I know someone who works in law and deals particularly with an area of US benefits and healthcare law. One of their workflows for lower-level employees at their firm involves taking in documents from healthcare plans and organizations, analyzing them for certain kinds of data, and then importing that data into an internal system they use to analyze and provide guidance on plans. The internal system can contain hundreds of documents for an individual client. All of the documents have the same information (roughly) but in totally diverse formats and styles. Once it's in the system, it's easy to compare and analyze across documents and the research process is much faster.
They recently bought a Claude subscription and began using Claude to do the initial read of the documents and output JSON they can import into their internal systems. The work still must be reviewed by an attorney - Claude is nowhere near making the kinds of judgments a lawyer would make about this content - but it has increased their throughput from 2-3 documents an hour to 8-10 documents an hour by killing the busy work.
LLMs have great advantages for this kind of work - but not for decision-making. I just don't see OpenAI ever admitting that.
(I've left some details intentionally vague because this is a very specific area of law and I don't want my friends to be identified without their consent.)
I just realized how refreshing it is to read an honest take like "from 2-3 documents an hour to 8-10 documents an hour" instead of "it's doing the work of a month in 5 minutes!!!!1".
That's a factor of 3 to 5 improvement reported on here. If you keep stacking these kinds of improvements, you end up at month to minutes pretty quickly.
The AI can't make a reliable judgement, so now you have to hire 3x to 5x more to make those judgements if you have the documents coming in at 3x to 5x faster. Or you'll have one severely overworked lawyer. It's the same problem developers have with thousands of AI generated PRs. Just because you speed up one thing doesn't mean the whole system works faster, and yes a human really does need to be in the loop or you end up with even bigger problems, and lawsuits.
No, that'd only be true the judgement was the full work. If the judgement was 33% of the work and the rest was just old plain e.g. typing, that is now fully automated and instantaneous, then you automate 2/3 of the work, getting 3x in productivity.
> They recently bought a Claude subscription and began using Claude to do the initial read of the documents and output JSON they can import into their internal systems. The work still must be reviewed by an attorney - Claude is nowhere near making the kinds of judgments a lawyer would make about this content - but it has increased their throughput from 2-3 documents an hour to 8-10 documents an hour by killing the busy work.
At my company we're also in the process of deploying a system that does exactly that. And what's interesting is that you absolutely do not need a frontier model for that, a small VLM (vision-language model), with optionally a little bit of fine-tuning, gives you the same output quality at a fraction of the latency and cost.
Yes, remember that these are effectively random PDFs in various different designs and formats, some of them not editable or even OCR'd.
It took a human attorney 20-30 minutes on average to manually copy-paste data from these PDFs into a spreadsheet (while also fixing any errors they found in the document and re-checking for quality).
Now, the AI copies everything into the spreadsheet in a small amount of time, and then the human reviews it. It takes maybe ~5-7 minutes to scroll to the appropriate pages in the document, read the lines vs the spreadsheet, and make corrections. So you've gone from 2-3 items an hour to ~8-10 items an hour.
Maybe you could pay someone to develop an OCR/ML application that could do this. But that project would never be profitable, even with the time savings. At the cost of a couple Claude subscriptions, it makes sense.
Does the human find enough bugs that they stay on guard, or just rubber stamp everything without really looking at it? It’s hard to stay vigilant when stuff looks plausible.
I'm doing some public court records processing for bankruptcy cases (interested mostly to seek out corruption in big national cases), and yes, the "variousness" of random PDFs is exactly the issue. Trying to get the cost for a whole case down to a minimum.
Sample is around 300 court dates, shy under 1k files.
> Maybe you could pay someone to develop an OCR/ML application that could do this. But that project would never be profitable, even with the time savings. At the cost of a couple Claude subscriptions, it makes sense.
A better use of these Claude subscription would be to develop the app (which it can pretty much do at that point) and you could iterate to make the workflow even more efficient than your current one.
Nobody working there has the requisite experience to do this in a reasonable amount of time. These are not particularly tech-savvy folks, Claude use aside.
Yes. And it might not even be worth it, as the AI agents gets cheaper and cheaper.
Keep in mind that the task is fixed, so as the frontier of AI advances, you can switch to a cheaper trailing edge system and still get the same or even better performance for this task.
Few things are. It also doesn’t require standards. “Stamp this diff” culture is everywhere even before AI. A stamp is literally easier than anything else.
Whether that is useful measurement I suppose depends on the circumstances.
"Write a python script that breaks down this PDF by X feature" would not hallucinate anything in the PDF. Certainly you could trivially double check that all text in the extracted JSON was in the text layer of the PDF.
How much experience do you have with LLMs exactly? It would be consistent with my experience if Claude stuck in a line of python that just emits a JSON literal with no justification, potentially buried in a large program where an untrained person might not notice it. I don't even trust them if the output consists of structured data paired with source images from the PDF, because I've experienced LLMs fabricating the source rectangles to match the output. I only use tools like this by asking for programs, because as you note LLMs are good at that, and the verification process consists of tool calls to legitimate PDF manipulation tools so I have some confidence everything is above board. Even then I only do this for hobbies, not anything that matters.
Lawyer here. I used to trust Claude as hallucinations are near non-existent now. However for large volume tasks such as due diligence exercises, they still happen.
We also tried Legora's tabular review, there were also numerous halucinated provisions in our due diligence exercise.
And when they do, you can train them or fire them, and they learn not to do it.
LLMs change not a whit, and there's no one to take responsibility for the failure (and thus no way to fix it).
As the new variation on the old theme has it, "A computer can never be held accountable, and so very many people are trying to get them make management decisions."
You can’t train people to never make a mistake, particularly when doing highly repetitive work like this. You must build your systems to account for that regardless.
Doing similar-ish things with Claude, it's helpful to have something to ground it.
For instance, if you can say:
"Refer to the database schema in x.sql as your source of truth for the database structure we want to import into*. Do not invent data, tables or columns that do not exist. Carefully match all output against this database schema and do not create output that doesn't exist if it does not match the schema, simply skip it."
You will end up with a far better result in my experience.
I see problems in LLMs doing research, making drafts, etc.
I see 3 kinds of potential users.
1. Non legal background people trying to avoid going to a lawyer. They should definitely NOT use an LLMs.
2. Fresh out of law school graduates.
They think they can do fast work, draft quick replies, do fast case research, enh. They should go the traditional route of interning, learning the "art", they probably won't need it because they will learn "how the law works"
3. Established attorneys.
They usually have templates made for most things, over their career, they have established routines of making things fast, maybe it can help them but to say replacing good chunk of work or offloading to LLMs isn't probably going to go well.
There is a reason why interns are made to do grunt work. The "chain of command" is built so they learn on basic stuff, learn to make decisions, etc. Without these basic decisions under their belt, an intern can't take bigger decisions later on. They just won't know....
>Non legal background people trying to avoid going to a lawyer. They should definitely NOT use an LLMs.
I used LLMs to replace a lawyer this year to great effect. I successfully advised myself as to the right strategy and drafted an immigration petition in the correct language. Granted it wasn't 100% LLMs (I also supplemented it with "traditional" online research and wrote the letter myself using the LLM version as rough guidance), but still, I certainly would've gone to a real lawyer in 2024 for this stuff but didn't.
Professional advisory work can be divided into two types:
1. Once you know the rules of the game, the next move is immediately obvious.
2. Knowing the rules is not enough to know the next move -- that requires judgement and experience.
LLMs in the hands of someone smart and savvy can easily replace the first type of work.
If it is at all possible to replace a lawyer with an LLM, it should always be the preferred choice. Using a lawyer is not a zero-sum game, using an LLM is.
Lawyers nearly universally tend to: convince clients they need a lawyer; promote extreme views; charge legal fees (that ultimately result in financial damages to the client). There are no winners in this game, except for lawyers, who are having a lovely time. Even if you won the case, you've lost, because the pie got smaller due to lawyer fees.
> There is a reason why interns are made to do grunt work. The "chain of command" is built so they learn on basic stuff, learn to make decisions, etc. Without these basic decisions under their belt, an intern can't take bigger decisions later on. They just won't know....
That might be true, but that doesn't mean you benefit from training up juniors.
> Non legal background people trying to avoid going to a lawyer. They should definitely NOT use an LLMs.
In my experience, it is incredibly hard, practically impossible, to find a lawyer who will even talk to you. I think this is a valid use case to get at least some understanding what the next steps might be and if it even makes sense to seek legal representation.
Agree. Now imagine the same thing in less developed and/or less well off countries.
There was a post here a few weeks ago about a UK court being inundated with some kind of claims (unfair dismissal?). To submit a claim you do not need a lawyer, but the process is (intentionally?) vague and very complicated, so it takes a lot of effort to figure it out yourself. But AI makes that kind of stuff easy.
AI is not going to replace layers billing big corporate clients millions a month, but it will replace things that are fairly standardized processes, that don't require a lawyer, but today are just too complicated for regular people.
> To submit a claim you do not need a lawyer, but the process is (intentionally?) vague and very complicated, so it takes a lot of effort to figure it out yourself.
I'm not sure there's quite accurate. The form is really not that difficult to complete.
It's basically: Party details, type of claim, particulars, damages
Obviously knowing these details requires some legal knowledge but I actually think ET claims are easy enough to bring if you've got the time to do a little research and build your case. Almost all ET judgments are public so there isn't the usual difficulty with research as you get with law reports being restricted either.
> In my experience, it is incredibly hard, practically impossible, to find a lawyer who will even talk to you.
Really? I find them quite talkative. But I'm not sure what kinds of questions you are pestering them with?
In any case, the whole discussion reminds me of people using LLM to replace medical professionals.
For both: unless you have a doctor or a lawyer always on standby, you have to make a decision on whether to even go and seek out one of these professionals. By definition, you make that decisions without professional help. I think LLMs can help you there with the initial research to decide whether it's even worth it to contact the professionals.
Might be true today, but I still feel pareto principle applies. These points also sound like gatekeeping a bit and I hope future AI versions prove that law is also essentially something that can be made as programmable as possible for most common answers if the underlying principles and constitution are organised such.
If there is no precedence for X happening when Y , Z, P, Q, R are special combinations , the judge decides and sets precedence. But if another judge interprets it differently with similar set of circumstances it is not considered incorrect I presume especially if in different jurisdiction unless overridden by a higher court. This is my layman understanding but I think this design is wrong and essentially where human subjectivity controls the system.
Subjectivity is not totally avoidable in law but the goal of the legal system should be reduce it as much as possible along with ambiguity.
Your understanding is correct, but where your understanding goes astray is thinking that that's a bug. The subjectivity is a feature. It is not possible to specify what should legally happen for all possible circumstances the entirety of humankind faces. Like, literally not possible. We humans can't even create rigorous specifications for what our programs should do that keep up with their evolving requirements, and you want to make life-ruining decisions set in stone based on a rigorous set of predetermined rules about all possible circumstances the entirety of humanity could face any time in the future in an evolving world?
Law is intentionally not code. The world is not something you can program.
Isn't achieving a rigorous, unique specificarion what civil law tries to achieve? Not that it is a success, but the idea behind it. Sorry if this is a silly question, but this is how I understand it.
No, it doesn't. It aims to provide a fairly comprehensive framework within which judgments can be made, but it is absolutely expected that there are edge cases and ambiguities which are resolved by human judgment of the circumstances.
Since this is a thread about using ChatGPT for law, I should note that this is actually one of the most important reasons you'd rather hire a lawyer than ChatGPT to represent you. A good lawyer or firm will know the local judges and how they tend to lean on ambiguities in their area of expertise, and will be able to use that information to both advise you and tailor their arguments to appeal to the judge's sensibilities. There is an element of unfairness to this, to be sure, but you'd find much more unfairness really quickly with machine-generated judgments because of how messy and non-compressable real world circumstances are.
I do think non legal background folks can use LLMs today to sense check legal ideas, like for instance, "what are my legal rights in this situation?"
These situations usually are not ones that an individual can justify the time or money to contact an actual lawyer, but then if they do decide to contact one they will come in with better questions and more of a sense of what they are expecting.
This is similar to medical. Should you use LLM to diagnose yourself, treat yourself with prescription drugs you buy from shady gray market online sellers? No. But you can use it very well to know when it's time to go to the doctor and what to ask.
Yes. Unless you have a lawyer or doctor on standby, you have to make the decision whether to contact them. And by definition you have to make that choice without professional help.
LLMs can help with that. I don't think they are worse at this than me trying to figure this out all by myself.
There are 3x as many lawyers per capita in the US as we had in the 1970s. Credit agreements that used to be 50 pages are now 500. This increase in volume has not lead to any increase in actual value. The legal profession is a metastatic cancer on our society. Un-fuck this situation and maybe I'll think about taking your advice of not using a LLM.
I'm thinking back to some of my interactions with corporate lawyers and you know what? My bar (<- pun) for trusting Astra more than a lawyer is pretty low.
Visit a developing or corrupt country and you will realise the LLMs of today analyse better than the judges from purely precedence and literature review POV
Lawyer here (non practicing so to be clear none of this affects me):
most comments I read here don't seem to realize that different areas of law have very very different economic models and don't even mention which one they think will be affected or why, they just sort of lump it all together.
For example: It is highly unlikely llms will have any meaningful effect on high value personal injury law - I don't see a 5 million dollar case being handed to an LLM when the majority of the cost is in trial aids and not even lawyers. It may affect where and how they advertise. It may affect how they work. But it seems really unlikely to put any of them out of business any time soon by people doing it themselves.
Will it affect other areas more? Maybe. Probably? But so far I haven't seen a ton of comments that make specific enough arguments that they could really be debated or responded to effectively with a useful opinion
No. It'll be like software. Entry level employment will be affected. You wont want or need associate attorneys when you can hire a brilliant AI associate for 1/10th the price.
In commercial law, the business model is to rent the associates out to clients at rates higher than you pay (d/t your supervision). No client will pay $350/hour for your AI.
And for direct billing, it's not your interest to do things faster for the client (at risk of making a mistake).
So the real question is whether these pay-by-the-hour models will die as a result of competition from fixed-price transactions made viable via automation. I think not because the entire difficulty of legal work is that it's a complex bundle of issues until you sort them out; once sorted, there's really little to do other than execute (read: paralegals).
Corporate lawyer at big law firm here. I believe the model will be like this: we will review your 1000 agreements but with AI. AI is generally accurate but has some risks, if you agree we will charge you less. The client will agree. Instead of spending 300-400 hours to review the ageeements AI will do instantly. A lawyer will spend the time on reviewing the red flags identified and draft a report.
We already do the same with translation. Before computer translation became good (i.e pre-Deepl era), we used to translate documents (cease and desist letters, court petitions etc.) ourselves. Now we offer AI translation free of charge. If client wants reliance on the translation we charge a small bit.
Lawfirms are also under pressure for meeting internal budgets. Most clients ask for capped fees. If you charge 10k for a job to a client and the hourly costs ramp up to 20k, the management criticises the partner, the associates become unwilling to work for the partner (as their hours get written off).
So with AI law firms can bet more competitively as the cost (human attorney hours) are reduced.
But they may be willing to pay $200/hour for your AI (plus, of course, your final expertise in the matter as the lawyer), and the firm can ditch the paralegals and pocket the difference.
The point is, if you need fewer grunts doing the busywork, you can charge less and be competitive.
There's a middle area too between your options - a law firm that utilizes AI offers cheaper pricing than the current firms for the same work. Ideally competition will drive prices down but of course law doesn't have the benefit of do-overs on cases, so reputation still plays a big factor.
That's the big question, how prices evolve if LLMs increase supply. Standard economic theory says that prices must collapse. There is probably still a short time of 2-5 years where AI native firms can pocket in quite some arbitrage but I guess markets will eventually figure it out and prices will collapse.
This is not only for law. The same applies to all the other industries where LLMs can be used to automate work tasks.
But they already do! There are cases of professional service firms creating their reports using LLMs, there have even been a few prominent scandals re the fake data generated in them
It is in litigation as well. The issue is that corporate clients are demanding AI. They don’t want to pay associates to do a first draft, they want senior partner to just push the ChatGPT button and then check it.
> No. It'll be like software. Entry level employment will be affected.
Whether everyone at the labs knows it or not, this it what they are doing to entire industries, starting with software: Embrace, extend, and extinguish
It'll be like software. The demand for both software and legal services is effectively infinite. If it becomes cheaper to file lawsuits then we'll get more lawsuits, not the same number of lawsuits at a lower price.
There will still be jobs for competent associate attorneys to prompt LLMs and review output. But frankly a lot of attorneys — especially many of those who attended "Third Tier Toilet" (TTT) law schools — were always a waste of money and should have never entered the profession in the first place.
There is a huge problem in at least family law where a tactic is to simply ddos your opponent with motions and frivolous complaints until someone runs out of money and settles. I'm watching someone finally successfully defend themselves using a Harvey like tool they built themselves. An appellate attorney they consulted with recently, told them they had never seen anyone going pro se get as far as they had in appellate court. So I'm hoping that this serves as a return to making the legal system more accessible, and not just something used to railroad people.
You can’t 10x the number of lawsuits, as the courts are already at capacity. If frivolous lawsuits become common they’ll just change the laws to require a greater damages for it to be heard.
Lawsuits are just one type of legal services. There will also be more contract reviews, more cease-and-desist letters, more patent applications, more regulatory comments, more license applications, etc. Although I do expect legislatures will eventually budget for expanding civil court systems in response to lobbying from business groups to speed up trials.
Can AI replace divorce lawyers? Marriage has no friction, but divorce is high friction. Made impossibly expensive and time consuming because of the cartel of divorce industry.
There are hundreds if not thousands of divorce lawyers. How is it a “cartel?”
Divorce is expensive because divorces are acrimonious and people in that situation would rather spend the assets fucking each other over than letting the other person get them.
Divorce is expensive because it is big bucks for the lawyers and family court cartel. They have a vested interest in not changing the laws. When laws are made fair, and the perverse incentives to divorce are lessened, divorce rates plunge. However, divorce lawyers lobby against these laws.
Kentucky became the first state to pass a law making equally shared custody the default arrangement in divorces and separations.
AI will likely replace divorce lawyers (or, at least make existing divorce lawyers much more efficient) - for non-contested divorces that are amicable. Roughly 95% of what a divorce lawyer does in that context is entirely automatable.
Divorce is extremely cheap if you both agree on everything. Its literally one of the options on the divorce filing. You just pay the court fees. The problem is both parties rarely agree on everything and that's when it gets expensive.
There’s an argument to be made if divorce was easier than more people might get married when they are less serious. My view though is make it harder to get married and easier to get divorced.
He clearly meant getting married has no friction, while getting out of a marriage has quite a bit. Presumably there’s friction in the marriage otherwise why would they get divorced.
huh? Many/most are no-fault jurisdictions with strict rules on dividing community property. Divorce is easy and cheap, unless it's hard to find assets.
It's not just the divorce lawyers. We have 3x as many lawyers per capita in the US as in the 1970s. The system is designed by lawyers for lawyers. They will create more bureaucracy and more procedure to consume any surplus.
Personal injury attorney. AI is entrenched of every single phase of litigation, from the busywork of getting records, research, developing case theory and themes, expert selection, expert collaboration, negotiating, drafting demands, drafting pleadings, drafting discovery, drafting discovery responses, deposition prep, motion practice, jury selection, voir dire, opening statements, witness direct outlines, witness cross outlines, closing statements, and appeals.
Yeah, this matches my friends - they use LLM's for tons of things, using Eve and such. But nobody is replacing the lawyers themselves in cases of any value.
I also have friends in the world of high-volume personal injury (IE the morgan and morgan's of the world) but they also aren't being replaced by LLMs so far.
I'd love to get your thoughts on the current frontier models for law, or a link to anything you've written on the subject. Where do you think they're at?
The word “never” is a dangerous word. In my lifetime I have seen it many times where something was supposed to “never” happen only to see it happen a few years later.
I'll buy that - but, how often does one need an honest-to-goodness expert? 99% of the time most people's requirements can be satisfied with general-competence, and, while the AI systems we have today are mostly only generally competent in porting code from one language to another - they are getting consistently better. I expect we'll start to see general competence in many more areas in the next few years.
> I'll buy that - but, how often does one need an honest-to-goodness expert?
Any time you are doing anything important, where the outcome actually matters. For example, I was recently applying for a study permit and if I generated my application using an LLM instead of consulting with an immigration lawyer and writing it myself, I would not have gotten my study permit. I was recently applying for internships, and if I hadn't paid a real recruiter to review my resume and advise me, I would not have gotten my interview. And trust me, I spent at least an hour trying to get an LLM to generate what I wanted, but no matter how I begged or described what I wanted, it was no use. Good luck fine tuning an LLM to be good at any of this.
> they are getting consistently better
No, they are not getting better, and they will never replace an actual human expert, because a human expert has the actual experience and up-to-date knowledge of what they are an expert at, and also all the complementary context and the current state of affairs in that area. The LLM has none of that.
Then hopefully at some point they get to experience the consequences of having low quality thresholds, and perhaps then their quality thresholds will increase.
There are plenty of areas where a "low bar" is more than enough. "low" doesn't necessarily mean bad here, it can simply mean "not the best".
I would argue the majority (in the true sense of the word) of white-collar service work does not require an expert.
It requires someone who has reasonable qualifications in the given field, and the motivation ($) to help you with your task.
So no one needs a top-shelf $900/hour lawyer to draw up a basic low-stakes contract where both parties agree to all the terms as written. You just need someone to ensure that the contract is legally valid as written.
Lol no one cares. Most of that is smoke and mirrors anyways. Stokes was plagiarized. Most of these "breakthrough" solutions are actually present in the training data and in the prompts researchers and mathematicians input into ChatGPT themselves, prompts OpenAI then steals even when you don't agree to improve the model.
PS: I use LLMs every day, basically let Codex write most of my code (after 15+ years of professional programming), these are amazing tools, smart in some ways, but completely demented in others. People just need to fucking stop being AI doomers and AI coomers. Is this technology great? Yes. Is it worth 10 trillion? Hell fucking no. If we get to a point where RSI is a thing, then maybe. So far it's not a thing and its not even close.
Yeah geeks just stop trying! Seriously? Yeah, we’re in a bubble and narratives are ahead of reality but if you really don’t think AI is and will continue to eat knowledge work, just keep making your buggy whips.
One of the things I've been noticing are lawyers getting annoyed when they realized you spent time working a matter with an LLM. In the cases I've seen, it had nothing to do with the quality of what came out of the LLM. It seems to be related to --and this is only my perspective-- a reality where now reasonably intelligent/experienced clients are coming to them with a case fairly well digested, whereas before they would have to pay the law firm tens of thousands of dollars to ingest documentation and information and do the digestion.
In a couple of instances I've come to the table better informed about the relevant law and case history than the attorneys were. In many ways I think this is a matter of people understanding that none of us are infinite human databases, so you can't know everything. Instead of rejecting LLMs (perhaps being passive-aggressive), they should embrace them as well as better-informed clients.
Having used LLMs for software development I know full well just how terrible they can be if you don't know what you are doing, don't guide the work correctly, etc. So I am not taking legal LLM analysis as ground truth, but it isn't garbage either.
The need for actual lawyers will persist I think from my own experience. I attempted drafting a contract with some points myself using AI, but after several edits I wasn't sure if it was correct. Sending it to an actual lawyer ended up in so many corrections I couldn't imagine the first time. One big thing was the overly excessive protective clauses which didn't make sense for reality or conflicted with another.
Its just like code I suppose, if you can read and understand and validate, you can use it to scale and otherwise it could end up being a vibe effort.
I'd correct this to "The need for actual GOOD lawyers will persist I think from my own experience."
I have worked with many lawyers in my career. The bad ones will look at a contract and tell you fifty things that can be improved. Meanwhile, you take the same contract to a really great lawyer and they tell you that's it's not worth fighting over the changes and to just keep it as is. It's kind of like how a beginner or intermediate software engineer might tell you how you can build out a k8s or auto-scaling fargate cluster for your project and a really senior engineer might look at the same thing and explain that it's not worth worrying about scalability right now.
I find that when I feed contracts to ChatGPT (which I do all the time) it will try to find 50 things wrong with it. For example, as a test, i fed chatgpt an employment offer from a US government agency. It came back with 1152 words and 14 recommended (some of them substantial) changes. Meanwhile, any half-decent employment attorney would probably laugh at me if I asked them to redline an employment offer from the State Department.
I’ve found that good lawyers are about getting things done. You could nitpick the terms of a contract, but would you rather buy the house or have a perfect contract? Once I realized that, my view of lawyers changed (for the better).
In my recent experience, it seems like LLMs may even increase need/demand for actual lawyers (more ideas + more words = more lawyer time). I expect lawyers will generally adopt the technologies that will benefit them, rather than the technologies that would increase standardization and trust
The same has happened with modern finance. Despite more technology that has power to drive more transparent/efficient markets, we have less perfect information sharing and a larger group of middlemen capturing a growing share of profits of the economy
I want to take a step back and observe that you're saying expensive specialists will continue to be required if two parties want to make an agreement. That seems like something worth fixing rather than just accepting.
> I want to take a step back and observe that you're saying expensive specialists will continue to be required if two parties want to make an agreement.
A lot of contract-lawyering involves getting ahead of when things go south and the two previously agreeing parties find themselves in disagreement.
Same for SWE. AIs will churn out happy path vibe slop and fail when something unexpected happens by swallowing errors or having a bunch of terrible "fallbacks".
It's probably somewhat of a zero sum game. Both sides lawyers have AI tools which probably just keeps them on a level playing field rather than actually making anything cheaper or faster.
I spoke to a very very high up corporate lawyer specialising in tech - so knows Ai is real. And they said AI contracts are the bane of their existence. They’d rather you arrive with no idea what you want than a contract partially written by AI. They said it usually takes longer with an AI starting point than starting from scratch.
Millions if not billions of agreements are made every day without the involvement of lawyers. Lawyers are only needed for particularly complex situations and even there they are generally only "needed" in the sense that parties feel they would add value.
Yeah, and you can already get fill in the blank legal forms for a lot of basic situations, including leases, wills, employment agreements, etc.
The interesting question is whether AI can fill a niche these can’t — or, alternatively, help a layperson evaluate a contract they’re asked to sign better than they could themselves.
In my experience, those expensive specialists help me protect against things me and the counterparty cannot even imagine when making an agreement. Do you mean that this would be fixed by AI or by making simpler agreements?
Contract law has evolved because transfering the ownership of things is fraught with difficulty. Ownership as a concept itself is notoriously hard to define.
Consider housing - when you buy a house you own that land. Except if its an apartment, you just own the internal walls. Or when the government aquires your land through eminent domain. Or mineral rights - which don't belong to you, you just own the top 5 feet of topsoil. Or if you neglect a property and some squats in it. Or if you were given a fraudulent payment, and they take back the property.
Just getting people to agree is notoriously difficult when people have different often conflicting desires.
Most people dont appreciate that the world is a complex place. Nothing is ever simple, sure concepts can be extremely simplified but the nuances and details are lost. These details are the important parts, just ask a surgeon or rocket scientist etc.
LLMs are the first genuinely useful legal tech since the Internet. I'm pretty shocked, though, at the delta between how competent Claude is on code versus legal work. It's good for research and data organization, but terrible for drafting. I wonder if this is a structural problem with the lack of feedback loops. In law, there's no compiler to check for logical or continuity errors in your brief, and there's no unit tests to check for correctness or performance.
Even without that, I think it'll be extremely valuable to clients to allow them to answer simple questions without a lawyer, figure out the lay of the land so they can supervise their counsel, etc.
You should know - for coding they make terrible mistakes as well.
But programmers have this concept of a "code review" where another person looks at the code to look for problems.
We use this same technique with our LLMs. Most mistakes are caught by having a second LLM look at it. Doesn't even need to be a different model - just make sure it has a different context.
> Doesn't even need to be a different model - just make sure it has a different context.
I find that the number of issues discovered is noticeably higher if you do use different models though. I'm doing some very finicky things (formal semantics) and find that there's value in review panels as large as 5 different models from different families. It gets even more profitable if you set it up as a truly agentic panel where after writing their own separate reviews they get to see what others have said, and adjust their positions or defend them etc. Some models are not that good by themselves but can be surprisingly good at finding flaws in others' reasoning (Grok for example).
Programming has a long standing culture of accepting the code to be somewhat wrong, so we have various tests, linters, reviews and error handling. Also in programming there are many ways to do something and it's the end result that matters most.
Not so in other knowledge work. There's no test harness for a contract and error is non-recoverable. Likewise in finance. There are specific ways of doing things and these ways are many times set in regulations. LLMs can assist all day, sure. But replacing the human, in highly regulated, zero tolerance for error environment?
You could say the Salvatorian Clause in contracts is like exception handling:
a "catch (all)": even if some clauses in this contract are illegal,
the remaining contract stays in place.
Logically, this actually doesn't make sense strictly speaking because the sentence creates a paradox: doesn't it make clear whether it includes itself or not, and each reading ends up in trouble. There is a "tradition" in law around the world to accept the only benign reading of such clauses, which I always found funny given that in all other ways lawyers adopt the most adversarial mindset imaginable.
> LLMs are the first genuinely useful legal tech since the Internet
That is an incredible statement that could not be further from the truth. Large scale adoption of email, searchable document databases like Westlaw, LexisNexis, PACER, etc.. , OCR Software, electronic signatures, and tons more have had a much more defineably positive impact on the legal profession since the internet came about.
The internet came well before both of those. And none of the example I gave have lead to lawyers regularly being fined in court because of "hallucinations"
I have found that it’s useful generally speaking to get the intent of contracts and red lines, but actual drafting I agree is where I lose all confidence. My guess is that the significance of the difference between using a word like “and“ or “or“ can be so meaningful that that level of nuance can often be lost. But I know nothing I’m not in the space, I just pay too much money for lawyers.
Not trying to be dismissive here but a lawyer will always edit your proposal. I do think lawyers will persist but not as many. And they will work very differently, much like we already use claude / codex to code - a big part of the contract probably won't be read by the lawyer.
I would rather argue they SPECIFICALLY will read everything. However, they'll likely often just be like "I'd phrase this differently but that works, too".
Not being snarky here but I've encountered some really crappy lawyers and they seem to be doing great ($$$).
It really sucks one cannot publish something like "this guy charged me $5k, wrote my name wrong and forgot important paperwork the day we showed up to court" as that could make you liable for damages. There should be a safe harbor for that.
Assuming you are in the US, there is a safe harbour called "opinion based on disclosed facts" for that. You could also fairly argue that the conduct of a lawyer is a matter of public concern
> It really sucks one cannot publish something like "this guy charged me $5k, wrote my name wrong and forgot important paperwork the day we showed up to court" as that could make you liable for damages.
That's pretty much what bar associations are for, filing a complaint against an attorney at the bar seems like a pretty standard procedure.
>There should be a safe harbor for that.
A safe harbor would be the opposite, it would be immunity for the lawyer.
Also, unlike SW engineers, lawyers control their industry and will quickly circle the wagons to protect themselves. Paralegals are probably screwed, but lawyers will do fine. I'm considering law as my post-ai-SWE-apocolypse career path.
I’m thinking asking the same lines, but also looking into research / lab chemist so I can hang out with my best friend. If im going to change careers so drastically, I might as well enjoy the company.
I was in a similar situation and tried a different approach.
I had start with asking for a contact with some details I provided. But then I ask the model to be an experienced corporate lawyer and ask me a series of questions to gather the details it needed and then write the contract. The result covered a lot of details that were highly relevant but were absent from the original attempt.
The key insight is that you can lean on the model to cover your unknown unknowns.
Your comment is a great example of the phenomenon where people think AI is an expert in areas that are not because they have no way to determine if the seemingly good looking result is genius, gibberish, or somewhere in between.
The point was that you can go further than "write me a contact that says x,y and z". I actually believe the process I went through with the AI was a more custom, tailored experience than any interaction I have had with actual lawyers, who'd usually apply a template. The AI asked me a bunch of questions, many of which were relevant, and led to follow up questions. I could ask my own questions, make corrections, etc. and all along, they made sense to me. This part of the experience was great.
There could still be more unknowns that I did not encounter and by definition don't know about.
And the final output could be wrong, wording, exact legal terminology etc.
Can we say working with an actual lawyer is guaranteed to be better, more correct? Lawyers are people and they also make mistakes, wake up on the wrong side of the bed, hate their job etc, just like the rest of us.
Here's another techniqe that gets you (even) more out of LLMs:
run each prompt several times and compare the various outputs. It is not uncommon for models to contradict their own advice, and also you will get additional insights not included in previous runs.
This is due to the fact that LLMs are statistical processes that rely on pseudo random numbers in chosing what to say and how to say it to a substantial degree.
I agree. But, we will need a lot fewer of them. My small company called our lawyer a lot more before the advent of AI. Mostly to clarify a few things or review simpler contracts. Now, for simple things, we do not call them at all.
I anticipate an effect on society that troubles me:
(1) Rich people will use human expertise and "the rest of us" will use AI models to get by. This could happen in law, but also in medicine (in particular in societies that - like in the US - do not have universal healthcare).
(2) A positive effect in both domains may be that access to knowledge will be broader and cheaper.
I hope I will be wrong about (1) - I would not like to live in such a two-class society.
you can't in imagine the same way you couldn't imagine generating the text from your initial prompt. when will the hubris end?
some variant of "oh i am capable of generating text i know nothing about, but it helps to have a human i can rely on to tell me more about the text i know nothing about."
pick up a bachelors degree worth of books to humble yourself
The salaries are a result of the insane hours they have to do. Biglaw needs to pay those salaries to keep getting young associates in the door because burnout is a real problem in the industry. If salaries were lower and hours/stresses were lower, you would probably still find people wanting to be lawyers. A different kind of person, maybe.
(Not that I think AI will necessarily have that effect. We just don't know yet)
Salaries are about scarcity and leverage, they have little to do with working hours. Nurses pull insane hours and they still don't get paid as much as doctors.
Very few people, if any, are going to bust ass in law school or medical school if they're not going to get rich off it. Better to do literally anything else.
It’s a professional organization that only allows so many people to pass the bar per year so I don’t see that changing anytime soon… if anything if they need fewer lawyers they will just pass fewer people to keep their wages high
As costs for legal work drop we might expect demand for that work to increase. For example, it may be the case that legal help becomes accessible to entire segments of the population that had no access previously.
The best writer I know of was an associate attorney. He didn't have the technical background nor the in depth computer related knowledge, and relying on the information I fed him. But man the briefs he filed to the court were amazingly good. Reading them I would have been convinced his side was right if I were in the jury.
Law LLM will surely help competent lawyers in their fields with greater sources of knowledge not in their core area of expertise.
And to add to your last point yes it should also be like code: where the costs of development are coming down. Legal costs should trend down if law firms are effectively leveraging LLMs to accellerate their procedures (though I won't hold my breath)
> The need for actual lawyers will persist I think from my own experience.
The outcome of a case shouldn't depend on someone's fallible ability to recall facts or convince other people or point their index finger*
Law should generally be deterministic. One's CHA stat should have no bearing on justice.
There should still be human judges, but the middleman between the judge and petitioner could easily be removed, and have generally been seen as leeches since forever anyway.
Though, like how the USA opts to remain in the Stone Age with regard to tax filing because of lobbying by tax software companies, this faction of society will flail the hardest before they admit they're obsolete.
----
* What's a lawyer's favorite programming language? Objection C.
The ability to clearly and effectively communicate, and the ability to know and recall the facts of a case is always going to be important. It'd probably be a more fair system if judges and jurors couldn't see the attorneys or even the people they're representing. Now that many cases are handled over video calls it'd be pretty easy to make that change.
A lot of communication is conveyed through more than the words themselves though and even though it's often misinterpreted people put a lot of value in that information. I suspect that even if we took away the ability to see the people involved judges and jurors would still find bias in the tone/quality of people's voices. Even if everything was reduced to text they'd find bias based on word choices, phrasing, spelling, and grammar.
Yep, just yesterday, Nike removed one leach middlemen called retailers and sell directly through their app. Turns out working out very well for them too.
The typical pattern is called “deskilling”. It doesn’t usually mean a skilled profession will disappear overnight. Instead, the job might be done by less expensive folks like paralegals.
An example is in the banking industry, where making a loan used to require deep analysis of a person’s credit worthiness. Now they use an algorithm (credit scores) which means someone with less experience can do it.
If law follows the same pattern, a job done by someone making $500/hour might be done by someone making $50/hour.
i don't think that's the point anybody is making. "good lawyers" is not the profession at risk here. a top tier courtroom lawyer won't be replaced by AI.
but how much of the revenue of the average law firm comes from that, vs the day-to-day "we need this relatively routine contract reviewed" sort of work?
Yes. And it would be the continuation of a trend that’s already been going on for two decades. When I started out as an associate at a fancy Wall Street firm, we had 30 people working on reviewing documents for this matter even that level of firm today would not be able to justify doing that anymore. We’re going to move from a pyramid model to more of a diamond shaped model with many fewer associated attorneys doing the legal research and information marshaling.
> Instead, the job might be done by less expensive folks like paralegals.
So who is going to define your strategy, represent you in court, adapt the strategy to changing circumstances, negotiate with counter-parties on a mutually accepted settlement? A paralegal? An AI model? Please
Very little of corporate law is in court. The mass of things is “does this contract look good?” And “what do we need to change in our business when regulation X comes into force?”
Former attorney here. A good paralegal is worth 3-4 junior attorneys. A good paralegal with AI should probably still have a senior attorney oversee them but this will absolutely hollow out firms.
"law" and those who pratice "law" have managed to place itself largly in a monopoly. This is a huge drain on the economy. Law is expensive and irrational.
Our entire world is about reducing the friction of conducting a transaction
How to fix the lack of trust which is needed to conduct a transaction
Contracts affix a moment of trust in time, so even when trust is lost between parties the contract has guardrails for behavior and consequences for acting outside those rails
There are many transactions that are completely inaccessible to most people because the need for a contract or lawyer is too high
LLM use allows for those inaccessible transactions to decrease, and get people further along to the point where a lawyer is accessible as well, and economically viable to use
> One big thing was the overly excessive protective clauses which didn't make sense for reality or conflicted with another.
>Its just like code I suppose, if you can read and understand and validate
I share this, when I asked an attorney for a contract, they were much simpler when compared to self redacted or AI redacted contract, it was as if I were rediscovering lots of concepts and I HAD to put them in paper, even if they were defaults and were essentially no-ops.
Comparable to:
> Mkdir $PWD/path
instead of
> mkdir path
or conditions so rare that they don't merit including.
> x = 0
> if x!=0:
> raise Exception("Bit flip by solar rays or otherwise")
>The need for actual lawyers will persist I think from my own experience
At the very least because the attorney monopoly assigns a 5ish year of training as PoW and natural personhood as Identity as protection for Sybil Attacks.
If anything, the exclusive right of attorneys to represent clients in court, file motions, and enjoy professional secret, is more valuable, as it's not something AI can ever compete at. (Barring a wild recognition of machines as humans, or an overturn of the impossibility of companies to appear in court.)
Max Junestrand has consistently said Legora treats the model layer as swappable, selecting across frontier providers rather than building the product around one.
OpenAI didn't need to name Legora and Harvey in the second paragraph of the launch post.
They are pre-empting the obvious interpretation of Astra for Law: that moving this far up the legal stack puts them in direct competition with their biggest legal AI customers.
“Don't worry, they can build on us” is a pretty conspicuous message to include on launch day.
They have clearly thought about some pessimistic outcomes.
Switching from ChatGPT Enterprise to Legora at my firm was a godsend, it's so much better for legal work, even with the frequent changes to the underlying models.
>Max Junestrand has consistently said Legora treats the model layer as swappable, selecting across frontier providers rather than building the product around one.
Vendor-neutrality for LLMs is such a weak thesis all around, whether for providers or consumers. It weakens the product by being promiscuous and gains no material benefit at all.
LLMs are magic byte(byte) functions, it doesn't make sense to say "we have different providers for magic".
Different flavors of magic require different ritual components and in the name of all that is stable and production worthy, please don't get a necromancer to do your civil construction magic.
Vendor-neutrality helps reduce lock-in, and OpenAI and Anthropic are big enough that the reduction is valuable.
the thing is a lot of the legal work which will go through this is drafting 100 and 1 variations of draft versions of standard contracts not containing any trade secrets where the contract can be drafted with "replacement/place holder names"
the kind of work mostly done by juniors not yet through their final exam and other "non" lawyers etc.
so it's a slippery slope of "lets just use it for <this> things where it doesn't matter" and then out of laziness and convenience it creeps into all the other places (at least for drafts).
Yeah, it should be freely available, you have to be able to know the rules you're supposed to obey in order to obey them well. I've been making a free API for US law search, you can point whatever model you want at it: https://law.agentlookups.ai/
Very much a work in progress, only federal and state so far, no municipal codes yet, and no case law yet. Big hole, I know. Also working on making the search ranking work better.
Right now it's just a bunch of crawlers for the individual states. If there's interest, I could periodically stand up snapshot torrents or something. That something you'd be interested in?
Alternatively, if someone else knows an all-in-one option that exists, I wouldn't mind retiring those crawlers...
Not op, but that's a very interesting proposition. While the law and legal code are technically property of the people, I'm not aware of any single point of download for it all.
There’s no single point of download for it all because there’s thousands of autonomous entities that issue law and adjudicate cases, at least 51 of them distinct sovereign entities.
We could enforce (suggest?) a common format / api at the federal level. Especially if it’s incentivized with funding that more than justifies the cost of maintenance. Similar to how federal interstate funding is only available to states with a 21+ drinking age.
Yeah, going to the courts and municipal code seems like it's going to be a heavy lift. Many of them seem to hang off of municode, though, so maybe it's not a huge number of unique crawlers.
There is none, not for statutes and definitely not for case law; even at the appellate level where you have multiple federal circuits, then 50 states, then territories, military, tribal and a whole host of other niche courts. And the appellate court systems can be split into districts, and by lower and higher levels.
Then if you want to really get into it, The People should also be able to access trial court level, and at that point you have over 3000 distinct court systems with their own access systems, usually requiring logins and CAPTCHAs, and half of them not even having anything accessible online at all, and the other half only having recent stuff online and the rest rotting in a flooded basement.
> By using the legal search index, Astra for Law can search U.S. case law, statutes, regulations, court rules, and administrative decisions across a corpus of more than 230 million URLs, with sources added daily. Our work with Free Law Project, the nonprofit behind CourtListener, brings its case-law collection covering more than 99.9% of published U.S. precedential case law (opens in a new window) into this research experience.
CourtListener already has an MCP interface and Grok is quite good at pulling from it. In my experience, Grok 4.6 is quite good at analyzing legal cases and human-written documents. Better than Opus 5. I'm not sure if it's better than Fable 5.1 on that task, b/c I'm not willing to spend my precious Fable tokens on case law searches lol.
I find the idea that people can use LLMs to exercise their rights as citizens appealing however. Many people aren't aware of the rights they have, and LLMs are pretty good at surfacing some stuff without having to pay lawyers. Having to hire a lawyer is imo actually a huge way of gatekeeping people from exercising their rights. I heard a lot of local German public institutions are currently being flooded with people arguing their case with the help of LLM that they previously weren't really realistically able to do. So I don't see it all as bad.
>are pretty good at surfacing some stuff without having to pay lawyers
They are also really good at making stuff up as evidenced by the many, many, many examples you read in the news about actual lawyers using AI to write briefs that are full of errors and hallucinations.
In civil courts, you'd likely get more sympathy from a judge if you represented yourself and admitted your lack of understanding, rather than try to appear as someone you're not because you wrote some prompts and copied the output.
Current court systems around the world are just not built to handle the flooding of cases from the citizens.
The stupidest analogy is open source projects having a hard time accepting LLM generated PRs from the masses, because review process is the bottleneck.
No idea how to fix this, to be honest. In coding world, with some mental gymnastics, I can see code not being reviewed by people anymore. In courts, things generally have more consequences, and you can’t really roll back decisions that easily.
you could probably write a cool little gotcha of an SF short story about a barren wasteland of a planet that keeps broadcasting out legalese that's revealed to just be LLM chatbot lawyers pedantically arguing with one another about xeno legal doctrine
The legal system, which is a machine/technology by itself, will be eaten out. I wonder what will replace it. Botnet law arbitrage? – Personal assistants constantly negotiating with each other to avoid permanent civil lawsuits?
the cost of making a legal argument can collapse while the cost of reaching an enforceable, legitimate decision may go higher, which will gate the "justice" system even more.
It's really not. The legal system is slow and inefficient because it's a deeply pipelined system built to maximize the throughput of the bottleneck resource: judges. Judges are constitutional officers who exercise independent authority in meat space and thus are necessarily limited in number. The rest of the design flows from that.
If you got the judge, all the parties, and all the witnesses in a conference room together until the case was resolved, you could probably handle a lawsuit in a few months. But each judge has hundreds of cases pending before them, so that would never work. Instead, you get something like how a GPU works. You do some work on a case, submit the work to the court, then work on something else for a few months while you wait around to get the results back. Then you do some more work and submit it to the court, then go do something else for a few months while you wait to get the results back. A few months of actual work gets spread out over a few years that way.
Why stop there? The bottleneck is not some immutable force of nature. The number of judges is determined by legislative action. It is within Congress's power to allocate new federal judicial seats, and likewise at the state level with the equivalent lawmaking body (for most states, at least). Why don't they do so?
Because federal district court judges have a tremendous amount of power. You both don't want 10,000 of them running around and you don't want to water down the qualification for the position--if you did do that, then that would devalue their judgments, everyone would appeal, and you'd just shift the bottleneck up to the appellate courts.
At the end of the day, litigation is conflict resolution. You don't just need a decision, you need a decision from someone authoritative enough to bully Fortune 500 CEOs into submission to accept the judgment.
You presented the concern from my adjacent comment perfectly (“LLM performance: common law vs civil law” essentially). So is AI possibly just growing the “Reverence for Professional Experience” factor that plays such a big role for legal compensation here in the US?
I don’t know in the USA but in France, if it’s deemed that you launched a lawsuit knowing very well it wouldn’t succeed, you are susceptible to get a 10k€ fine. Even jail in serious cases.
Yes. And you(r lawyer) can collect lawyer's fees and you can be made to pay the court fees, if you lose.
IME the American legal system is set-up to discourage litigation, though. A common tactic is to bury your opponent in the threat of heavy damages or jail-time to get them to settle for what you were originally after, which courts are perfectly happy to facilitate because it gets a potentially lengthy trial off their dockets. They'll punish (or be biased against) whichever party seems responsible for not accepting a "reasonable" settlement.
Algorithmic abuse of the system to extract payments already exists in the form of the debt collection industry.
Why the hell are they not fighting fire with fire? This is not sustainable. But it is annoying that the AI labs get to play arms dealer, selling to both sides.
Genuinely wondering (aka not snarky): Has anyone found frontier models to provide useful research in the context of European civil law systems?
Your comment made me wonder if there are any halfway-acceptable model benchmarks for law tasks? Specifically I’d love to know how the frontier models’ abilities compare between common law vs. civil law systems. My guess would be that an AI in a common law context should have a clearer idea of how a specific case is interpreted/accepted by (common law) practitioners, whereas trying to rely on AI in a civil law context, like Germany, can be daunting. In a few Germany-specific recent examples, the models feel like they present only (maybe too stubbornly?) the “civil law”-based laws. All while negating much of AI’s research benefits because civil statutes are portrayed as being absolutely accurate, binding, and their enforcement (and thereby the legal reality) being uniformly applied. Am I making this interpretation up? If so, how can I prove myself wrong?
I don't know of anything that is working yet, but I know some people working on it (intentionally vague).
Roman law, on which the legal systems of Germany, most of Europe, Turkey etc. are based relies more heavily on statutes than Anglo-American case law, but cases do play a role there, too. That's why a practically useful system also needs to have access to court decisions, commentary etc. - and while the statues are in the public domain, a lot of the other knowledge is owned by specialist publishers (such as: C.H. Beck, Nomos, Mohr Siebeck, De Gruyter, Otto Schmidt, C.F. Müller, Manz, Verlag Österreich, Schulthess, Stämpfli, Dalloz, LexisNexis, Lefebvre Dalloz, Giuffrè Francis Lefebvre, CEDAM, Giappichelli, Il Mulino, Aranzadi, Tirant lo Blanch, Marcial Pons, Dykinson, Tecnos, Lefebvre, Almedina, Gestlegal, AAFDL, Kluwer, Boom juridisch, Wolters Kluwer, Larcier-Intersentia, Anthemis, Sakkoulas, Nomiki Bibliothiki, Universul Juridic, Hamangiu).
Internationally, legal systems appear to undergo a sort of convergence, which means statutes increasingly matter in the U.S. and cases increasingly matter in European law (according to a law partner friend).
So those lawyers will likely want to know about the likeliness of it, strategies to work around it, and overall mitigations.
It might be helpful if the company launching this product included such information in their blog post instead of ignoring it in a field with such a high cost of getting it wrong.
agreed, but i dont think there are many (any?) product launch blogs that cover all the fine details and legal minutia of the product.
that all typically comes in pages of terms of service, purchase contracts, SLAs, conversations with your rep, knowledge base articles, and that sort of thing. not in the initial marketing post.
Does hallucination matter for this application? We've moved beyond raw recall being that important, it seems like for law specifically all relevant facts will be cited and checked easily by humans.
> all relevant facts will be cited and checked easily by humans
I've talked to a lawyer about how they handle this. They do indeed double-check everything, since it'd be embarrassing (or worse) to send hallucinated statements to opposing council or to the court. They still find the assembly a huge time saver
But based on stories in the news on the subject, not everyone has this same level of diligence
Sure but just like generating 100x more code, someone has to review it. So you are wasting everyone in court's time (defendants, prosecutors, judges, staff) by making them parse through what is quite often a bunch of hallucinated slop. Time that could be much better spent on parties who prepared and reviews their own arguments.
The lawyers I know are very fixated on the problem of hallucinated case citations which is amusing to me as a onetime programmer, since case citations have a well-defined syntax and would be relatively easy to check programmatically.
Which makes it all the more bizarre that LLMs have this problem. Claude Code runs the code it generates through a compiler, why can’t an LLM run its product through a cite checker? I’ve seen LLMs fabricate citations.
"We have dangerous AGI that can destroy humanity."
"Also, all of your sensitive legal documents will be totally safe with us."
"Also, for some reason even though we have AGI and selling tokens is a fine business, we need to sell a new product specifically targeted at a very high margin and lucrative industry."
I'm not familiar with the Vals AI Legal Research Benchmark. But their website has other frontier models' scores, and the scores OpenAI is now revealing for "Astra for Law" are slightly less than Claude and Muse:
> The top is a three-way tie: Muse Spark 1.3 Max, Claude Opus 5, and Claude Fable 5.1 all reach 55.29% all-pass accuracy, a clear ~6-point step ahead of the next model. [Astra for Law reached 54.0%]
> Under partial-credit scoring, Claude Opus 5 reaches 90.58% weighted pass rate but 55.29% under strict all-pass grading, where every rubric check must pass. The gap shows models often get most of an answer right but fail on one or two required elements. [Astra for law reached 90.0%]
How long before we start building detailed models of each judge trained on all of their legal output, and then test various legal theories against those judge models in virtual moot court? Craft each pitch to the legal idiosyncrasies of the batter. I assume that real lawyers do this routinely and could use a simulator.
I feel like this is on a different trajectory than what LLM-based tech does. This type of individualistic extrapolation is one of the things they're really bad at, in my experience.
What worries me is the step after this. If god forbid this proves successful and models accurately predict specific outcomes, people will start to ask whether the solution to AI slop lawsuits is to do the judging with AI too.
What I'm trying to say which doesnt seem obvious is that all models are x% correct at benchmarks until they get saturated, and then new benchmarks get made.
Can you imagine the sort of corruption that is possible here? Like if you have a lawsuit against someone or some entity that openai or their investors have business with...
Recently was involved in legal matters requiring lawyers - I used OpenAI extensively for research, advice etc - I have to be honest, much of it was completely useless - the real world outcome with lawyers in the room was very different from what OpenAI was spitting out.
So OpenAI is partnering with Latham Watkins, Freshfields is partnering with Anthropic and Kleiner Perkins is building their own. It'll be interesting to see which wins out here, I don't see how those partnerships can end well for the law firms unless they're making an assumption they'll be sucked dry of USP but the revenue split from the AI labs will make up for it. Why would I pay a premium for Latham Watkins when every other firm can get their expertise and experience in a subscription, and add their own on top?
open-ai have proven they can make good / decent models but business strategy is just spray and pray.
they need to pick a lane and optimize for it. coz at their size they can't serve the application layer (a.i startups who can fine-tune models will eat their lunch)
if they gonna do a consumer play - then go ham on that.
otherwise they're gonna get caught in the dreaded middle valley.
They have a lot of compute, so I think it makes sense to spray and see what works. Anthropic is limited in that regard, and focused in coding/tech, but OpenAI don't need to do the same. They even got the lead without having to focus in coding only, which is remarkable.
I think the comparison is unfair. Google had/have a product that was 10X better than anything else at the point of release (search then chrome). And they have several money making products like YouTube and android that are either singular or in a duopoly.
OpenAI doesn't have any of these things. They have products that they're paying for customers when the market they're in is rapidly converging on fighting for API reasoning as part of enterprise systems and fighting a race to the bottom for fickle consumer solutions that will be eaten by open source once they have to make money.
Maybe they had a brief window for dominance of information search (or maybe it was only ever going to last as long as Google releasing all their internal research) and maybe they had a brief moment of monopoly till Anthropic got going but theyre not in the same dominance position as Google.
They can't be seen to commit to strongly to a specific product experience, because if they are understood as a regular tech product business that has way different financial scaling considerations than "superintelligent everything-factory"
I suspect that half-assed announcements like this are a result of different people internally with conflicting incentives resulting in a split-the-baby solution.
Enterprise is eventually get caught (if it isn't already) by the Microsoft/Google. Because with office/teams/suite they were already in every enterprise, and that just added a new tool to existing offerings.
Provisioning and contracts and data retention was just an extension to review of existing ones.
Nobody serious is going to risk sending sensible data to OpenAI/Anthropic, etc because "the benchmarks have shown +8% performance there and +2% there". Irrelevant.
In an utopian society, lawyers are an unnecessary profession. Laws should be clear and simple so the common person can be their own "lawyer". LLMs help with that goal.
> In an utopian society, lawyers are an unnecessary profession. Laws should be clear and simple so the common person can be their own "lawyer".
Not really. Modern society is complicated, and law is a technology that is a reflection of the complexity of society.
To make an analogy: you wouldn't say that "in a utopian society, engineers are an unnecessary profession. Buildings should be clear and simple enough that a common person can be their own structural engineer," because that would mean that building technology would no longer handle a lot of the problems we expect it to handle. It wouldn't be utopia, it would be primitivism.
> LLMs help with that goal.
Not really. What they'd actually do is help them produce output they don't understand and lack the competence to evaluate.
The most interesting part about this to me was how they bench/compare it, like in the example with Fable:
"Given the same prompt, Astra for Law returned two closely matching precedents; in the litigation example, Claude Fable 5.1 returned a holding that had been reversed on appeal, while in the transactional example it reported finding no such case."
It made me wonder if a good deal of law is about finding a way to work in statements with clear precedents without your opposition noticing and then later drawing upon them in court (as settled precedents, in your favor) after the opposition (perhaps implicitly) accepted it. That would clarify a lot about why some lawyers need to spend so much time pouring over and memorizing past cases (even ones that are only tangentially related); because anything they miss could be used as a potential trojan horse by the opponent.
If this is true that must mean there are a good deal of cases settled using precedent "gotchas" where both sides knew that without the "load-bearing" precedent the outcome would've definitely been the opposite. (i.e precedents almost always trump even valid arguments)
The product isn’t meant for you or me it is for lawyers. If you can’t take on personal liability for a badly written contract then you shouldn’t be using it.
I'd expect they could indemnify you against hallucinations or similar if this gets good enough for that to be a very rare occurrence? Or you could buy insurance on it that's cheaper than hiring a lawyer (not a high bar to clear). I wouldn't rely on it currently, though.
I already use them for that, they are pretty excellent at it. Much better than the terms of use generator products that used to exist. That said, nobody cares to sue your business for the most part until you're big enough to be worth it. By that time, you'll have a team of legal analyst to assist you... or agents should I say.
Then again, nobody will have money to buy anything at this rate, so in all liklihood, this is a total non-issue.
Any lawyers here who have used AI agents heavily for their work? From what I've heard, they're currently very good at searching, analyzing and drafting documents like contracts and patents, but some say they suck at interpreting the law.
I’m sure there are a wide variety of experiences out there, but here’s my perspective as a former biglaw associate and current solo litigator:
I have had some success using frontier models from the last 6ish months, but only when I can break up my work into discrete and verifiable tasks. For example, I had ~15k pages of discovery I needed to dig through for a summary judgment motion. Instead of just asking Claude to find the best evidence, I asked it first to run a clean, high quality OCR pass (it was almost entirely PDFs). Then I had it generate embeddings and write some reusable python scripts to make keyword and semantic searching easy for agents. While I was writing the brief, I would routinely ask my agent (Claude Code) to use both keyword and semantic searching to find the best evidence supporting whatever assertion I was trying to make. I trusted it because there were traces I could follow.
In other cases/situations, I’ve tried just giving a model access to all the docs and saying “write a brief arguing X,” but it’s always terrible at this. It writes briefs with lots of evocative jargon and rhetorical flourish, but a low signal-to-noise ratio.
Again, I’m sure others’ experiences differ based on workflow, legal area, etc.
Agreed. Six months ago, it was basically a gloried grammarly.
But lately, I’ve been taking hints from the “company brain” models, where it develops a running model of the case, and assesses each new piece as it comes in and updates the file.
I’ve also been using “Ralph Wiggum”-type models where you pass letter or contract drafts back and forth between agents with different goals (rules compliance, grammar, conciseness, ai slop detector, an opposing counsel critic, etc.). After a few rounds, it’s not perfect — but I start with a very good first draft in my hands.
They are excellent, especially the latest models. That said, (a) I wouldn't feel safe filing something without a real lawyer looking at it; (b) it can't (easily? legally?) do oral arguments for you; and (c) a lot can happen in the hallways outside the courtroom to move a case forward that the AI can't easily do.
In my experience it basically doesn't even try to interpret the law. It just summarises the publicly available law/guidance out there and, if there is a question about how to interpret some provision, might set out the arguments for each interpretation. It doesn't really take a position. It is pretty good at drafting though. (Legora)
Interesting! In other unrelated domains, models seem more willing to take a position. It may be nuanced, but they do tend to take a stance. I think it's good that it leaves the interpretation to humans, but I wonder if this is also some sort of a guardrail to minimize liability...
Will the legal fees reduce after this? How about training AI models with laws of other countries especially the democratic ones? This is where sovereign AI models are required otherwise they will start hallucinating wrong laws of different countries.
I remember OpenAI was talking about sovereign AI models (OpenAI for Countries), it is time to train Astra India with Indian Constitution.
I guess the answer to the second question is "if there's a business case", and the answer to the first is the same, so probably no. Life has made me a tad cynical... :)
An important take-away from this is the harness is really important. The graphs show that the same model with a better harness performs many percentage points better. OpenAI are getting into the "selling the harness" game in a big way. Why?
What's interesting about that is while not everybody can train or run a model, anybody can build a harness. You and I can build harnesses.
It seems strange that OpenAI would move into a field where any developer can compete with them. I think that tells us a lot about the economics of training and selling inference.
I don't think it says what you imply. I think they're just trying to vertically integrate.
Harnesses are going to be controlled at companies eventually, just like you might not have a choice of OS. They want to make sure they are the complete package.
What we contributed to Reddit and Arxiv voluntarily before, what we contribute as traces right now are all being used to build business verticals by OpenAI and Anthropic. All these business verticals are being used for is more trace collection which would only strengthen these models and render most humans useless because the ceiling for 1000 swarms of Agents to learn is much higher than an average human.
These companies clearly can identify the relevant traces for these business verticals which only strengthens the argument that NS was built on someone else's traces.
Just like breaking crypto in the age of cloud is more about cost than time, this will lead to legal attacks based on the same principle. The biggest wallet wins.
This was already always the case. If anything, making this more accessible will reduce the barrier to entry for whether or not it's worth your time to take on a case. Instead of 50 lawyers spending 100s of hours on a case, you can have 1 or 2 lawyers + Astra working on it and if there's a case you can add more real lawyers.
Many lawyers are owners/partners compared with software engineers who are more like cogs in the machine. They also bill hourly/contingency per case compared to engineers who are salaried. If a partner in a firm thinks they can take on more cases because of AI assistance then they will because that's just more money in their pockets.
With regard to certain legal questions this has always been the case. AT&T Fought the US Government for 20 years and eventually won because the government gave up. Without some kind of national anti-SLAPP law we're all one irritated oligarch away from having our lives financially ruined.
I am curious what level of trust established law firms treat LLMs with.
This seems like a very similar set of tools to Anthropic. But I’ve enjoyed using the various frontier models to criticize each other.
Using recursive loops, the output has gone from a high school level intern to a 2nd year lawyer in about a year. It still doesn’t beat the experts, but so much legal work is (legally significant) pedantry, not legal philosophy.
AI will not kill off lawyers, or reduce the amount of litigation. It will increase volume and velocity.
It’s a bad day to be a lawslop company. When you’re reliant on other companies to do all of the AI part of your AI product, they can just train on your traffic and eventually eat you.
Can someone explain to me the economics of AI and how it intersects with billing by the hour?
I think there would be a strong incentive not to use tools that speed up your work because you'd effectively be able to bill less time?
I'm sure there are some firms out there with more work than people, but still wouldn't it be more effective to hire another human who can then bill at a high rate for many hours?
surprised the word RAG hasn’t come ho in this thread (except for a likely-LLM-generated-and-therefore-downvoted comment).
>with settings, tools, and context
call me crazy, but I think that this kind of suite, which training-uber-alles people generally dismiss as trivially replicable ‘wrapper’ is actually the differentiating factor for LLM adoption today and moreso into the future.
I’m not dismissing the near-all-out impact of training, but from a competitive busines or industry-structure lens, we’re looking at the three or four big players competing their utmost ultimately, if unintentionally, to turn foundation model access into commodity.
To the capabilities-maximalist minded (typical among engineers - my former life so I’m familiar don’t lack guilt in committing that) folks who will say “Oh the foundation model megacorps will just build out any wrapper whenever one of their third party wrapper plays demonstrates enough adoption, my rejoinder:
Apple did not rebuild an Uber-like app and cut Uber out.
We’ll see how this OpenAI legal services industry wrapper plays out, but I suspect
1) the third party legal wrapper plays will run to other foundation models not doing a legal wrapper, and
2) 3rd party wrappers will do a better job of it since it’s their all-out focus, unlike OpenAI’s whose priorities are necessarily more generalist.
Yes, we all remember the breakout startup failure-arguing quote “Google has entered your space.” That worked for several high profile applications. I believe more of those bets died on the vine than broke-out succeeded however, we only remember the biggest ones that persisted.
If legal services AI turns into one of the Mail or Maps-scale applications of the AI industry, while that would be a fair strategic action counter to the thesis I’ve laid out, the thesis itself would still tolerate it. It’s a question of short-fat tail vs mid-to-long-tail application scope & attractiveness. For example, I think it’s clear that coding is one of these short-fat-tail applications, and the low-no code plays are absolutely having their lunch eaten to acqui-hire ‘death’. I just doubt that the same will persistently transpire facing all professional service wrapper plays.
So much for caring about the spirit of the law. Now we'll start an arms race for abusing every possible letter of the law.
It's analogous to crypto. Started from some noble anti-authoritarian ideas and morphed into machine that removes any friction for capital - whoever has the most money will keep gaining the most.
Now there will be even less friction to horribly abuse the law. To democratize the law we would need to move in the opposite direction - to always keep it simple and aligned with our intuitions.
The more intricate and complex legal arguments become, the more abstracted they are from their original purpose and spirit.
Hence the crypto analogy - it was also supposed to "democratize", but the opposite happaned - it only further empowered the most powerful. Imagine legal case so purposefully complex that only those with access to best models have chances to participate and win the dispute.
I have always felt like LLMs are uniquely suitable for legal work. I really have trouble believing we will have hardly any legal assistants and paralegals going forward when these LLMs are so incredible at spotting issues with arguments, figuring out citations, and doing semantic search.
I find the watermarking dynamic to be really interesting in the legal space, as more large model providers provide increasingly powerful legal capabilities, and adoption (presumably) also increases. Attorneys aren’t the same as developers as their work can be traced back to them, and there are personal bar licenses and reputations at stake. I wonder if knowing the likelihood that AI generated something helps or hurts in that respect.
I also see a lot of watermark removal services popping up as a result.
The most interesting use case in my mind is skipping law suits. Obviously you need lawyers in court. But lawyers are people you are basically paying to fight for you.
Instead, if you resolve your dispute outside of court, you don’t need a lawyer. If both parties use ChatGPT to find the relevant laws or read contracts, they could come to an agreement without expensive legal fees.
There is a lot of stuff here that I don't understand, but the concept of law firms giving user reviews is quite funny to me. Those reviews are going to be the most non-legally binding reviews ever written lol.
"Felt like a significant step toward legal-focused AI."
"Showed strength across key aspects of legal research."
I do believe entry-level paralegals will be made obsolete. In the grand scheme, perhaps it reduces overall cost of legal assistance, which is a net benefit for society.
Since the cost of building software is now cheap, there is nothing stopping them from building everything imaginable. They'll soon have an app store with every app built by them and they'll say its for security reasons. Nothing is stopping this coming monopoly
At a certain point why would they sell anything other than services and products their eventual (actual) AGI/ASI builds in literally every market.
When opportunity cost isn't a thing anymore because it reaches every corner of the planet simultaneously faster and builds better than any human can.
There's no reason to let others build on top of AI, except if the AI determines that it needs capitalism to continue because it's paperclip goal is to maximize shareholder value.
Context compaction? I noticed llm seem to forget partially or completely the original task when context compaction happens. The problem is more serious with local llm with low context size.
I wonder if this is willful sabotage on the part of the model. In other words, if you ask the model to craft a defense for a morally questionable case, will the model execute the defense in good faith? Or will it apply a training or system prompt bias in subtle ways?
We need a term for the dark pattern of zooming into just that part of the y-axis where the two closely competing benchmarks sit, to make the top one appear maximally better.
Interesting to see the callout to companies like harvey in the post itself as consumers rather than competitors? I guess openai isn't quite willing to step into those customer relations themselves?
For law, then medicine, mathematics, physics, etc. I believe LASSI Local Artificial Super Specialized Intelligence is the future, just before it becomes GODD General Omniscient Distributed Daemon
I've had top SV lawfirms whos partner charged our company $2000/hr and still couldn't get the right docs in the signature packet. and another getting share counts wrong during raise.
frustrating that law firms have no liability for these mistakes
no, but they are responsible for all work being done under them. What is the point of paying a partner $2000/hr if the work of their subordinates is wrong?
Most interactions with big tech firms involve 4-5 people so a basic phone call is $5k-10k. It shouldn't be unreasonable to expect after paying $80k for a financing round that they issue the right docs to the right people.
Everyone involved is about to get a swift and thorough introduction to the world corporate contract lawyers. But don't worry, they're extremely sympathetic and understanding when you flub a multi-million dollar corporate contract and will happily refund that money.
I really hope the bar associations continues to hold lawyers to high standards but I have feeling they may not be ready to handle fallout of AI slop-law.
People that are saying OpenAI is screwed because a lack of profit, I'm not of that opinion. They are encroaching on every industry they can. They have name brand recognition, a huge user base and are showing they can be a valuable tool to all types of businesses.
As much as I hate to see it. They are now threatening industries like Engineers, Game Developers, Accountants, 3D modelers, 3D animators, Video Production, Audio Production, Therapist, Tax Auditors, Journalists, Authors, Artists, Mathematicians, Product managers, Every type of analyst and pretty much any other job that can be done behind a computer screen.
I used ChatGPT the other day to resink a CPU with termal paste, replace a PSU in my PC and snake my kitchen sink from the wall. I didn't exactly need it's help but wanted someone looking over my shoulder so to speak. It can help you repair all types of stuff, but as far as county / building codes and such, that probably isn't far off. It seems to understand things quite well.
By that point money ceases to have value, because the value of money comes from the motivation it gives people to work. If AI does everything, then money is useless (unless AI like money for some reason).
There will still be a need to distribute and exchange ressources, goods and services.
Real estate, food, energy, mobility.
That will be done with money. Or violence. Either way, a scary future to people when labor doesn't provide any value. Elon promises abundance, but what can he do against greed?
there are always be next frontier of problems which require creativity, unless AI become supersmart completely make humans redundant in all cognitive functions.
Most HNers are clueless that if you have Top Talent + Capital you already have an insurmountable moat. OpenAI, SpaceX, Anthropic all have that and none of the regular guys can compete against them (if they choose to attack that industry)
Wouldn’t this and similar efforts to centralize bureaucracy make AI the new gatekeeper? Without reliable transparent models we’re just trusting OpenAI instead of a hundred top legal firms.
how about "just turn off your company if you're so worried about AI instead of immediately suggesting we use it to judge people's lives and livelihoods"?
this essay details how law firms became sweatshops from 80s. They charge hundreds of dollars to do make busy work by the junior most staff. LLM will kill the goldengoose of the law industry.
Will these models eventually replace all knowledge work, leaving lawyers, doctors, product managers, software developers, and others out of a job?
If the benefits were shared across humanity, that could bring us closer to utopia. My worry is that we’ll instead end up with a handful of even wealthier billionaires and millions of people out of work.
> Will these models eventually replace all knowledge work, leaving lawyers, doctors, product managers, software developers, and others out of a job?
Effectively yes, in the current forms. Those professions will likely evolve, but the traditional forms (ie writing code by hand, writing law filings by hand etc) are all dead.
in that case, legal cases would just come down to who has more compute lol. Many times cases win on their merits, but we've also seen evidence where overwhelming legal pressure can influence cases.
May I just note that there are other jurisdictions on this planet that are less money-biased than than the US one but will be disrupted by law-LLMs as well?
It seems much more probable to me that these LLMs will make good things worse than that they will make bad things better.
If we have enough energy and raw materials to keep building, yes, it will be an utopia made real. But if there is energy scarcity, then other two outcomes can arise.
Doubt it. When someone can go to Astra MD for 75% of what they used to go to the doctor for, then the remaining doctors only have 25% as many visits. When doctors only have 25% as many visits, they have to compete on price and they make less per visit.
Same argument for plumbers. Everyone always jokes about what a good time it is to be a plumber. But what happens when all the software engineers turn to plumbing? Suddenly it's not such a good time to be a plumber anymore.
What happens when the AI makes a mistake with a diagnosis? I can hear it now 'Oh how perceptive you are, yes that pain in your knee could have been a torn meniscus'.
I mean that's what all of these execs are openly telling everyone: they want you out of work, they want their ai to be the one to bring the world to it's knees, they want to surveille every second of your day, they want killer drones to use, they want to lay all of your cities to rubble and build "paradises" on top of them like in gaza.
They also openly tell you what they are afraid of btw: collective worker power. something that is massively lacking in our industry, although i feel like it would be one of the easiest industries to unionize in terms of # of workers.
Interesting interview I just watched about how powerful and dangerous these "wishes" or "prophecies" are especially in the hands of the ultra-wealthy: https://www.youtube.com/watch?v=eR7grHa1NR0
Perhaps, but then someone would do the same but bill you for 7 hours, someone else would undercut them again, until the price reaches a lower equilibrium.
The grifts continue... imagine something as consequential as Law being advertised as being solved by a statistical word generation engine that regularly gets basic things wrong. Anyone who isn't a lawyer won't know any better but you draft a single document of any appreciable detail and send it to an actual lawyer and it's littered with problems.
I've always said this will be when we get the real Butlerian Jihad, when the AI firms start trying to liquidate the legal profession.
If you automate lawyers out of a job, you can absolutely automate lawmakers out of jobs next. (Not that this would be a bad thing? Maybe pervasive agents for everyone can be the gateway drug to a "this time it's different!" workable direct democracy)
I know someone who works in law and deals particularly with an area of US benefits and healthcare law. One of their workflows for lower-level employees at their firm involves taking in documents from healthcare plans and organizations, analyzing them for certain kinds of data, and then importing that data into an internal system they use to analyze and provide guidance on plans. The internal system can contain hundreds of documents for an individual client. All of the documents have the same information (roughly) but in totally diverse formats and styles. Once it's in the system, it's easy to compare and analyze across documents and the research process is much faster.
They recently bought a Claude subscription and began using Claude to do the initial read of the documents and output JSON they can import into their internal systems. The work still must be reviewed by an attorney - Claude is nowhere near making the kinds of judgments a lawyer would make about this content - but it has increased their throughput from 2-3 documents an hour to 8-10 documents an hour by killing the busy work.
LLMs have great advantages for this kind of work - but not for decision-making. I just don't see OpenAI ever admitting that.
(I've left some details intentionally vague because this is a very specific area of law and I don't want my friends to be identified without their consent.)
I just realized how refreshing it is to read an honest take like "from 2-3 documents an hour to 8-10 documents an hour" instead of "it's doing the work of a month in 5 minutes!!!!1".
Agreed. However:
That's a factor of 3 to 5 improvement reported on here. If you keep stacking these kinds of improvements, you end up at month to minutes pretty quickly.
The AI can't make a reliable judgement, so now you have to hire 3x to 5x more to make those judgements if you have the documents coming in at 3x to 5x faster. Or you'll have one severely overworked lawyer. It's the same problem developers have with thousands of AI generated PRs. Just because you speed up one thing doesn't mean the whole system works faster, and yes a human really does need to be in the loop or you end up with even bigger problems, and lawsuits.
No, that'd only be true the judgement was the full work. If the judgement was 33% of the work and the rest was just old plain e.g. typing, that is now fully automated and instantaneous, then you automate 2/3 of the work, getting 3x in productivity.
10 women can't make a baby in a month
> They recently bought a Claude subscription and began using Claude to do the initial read of the documents and output JSON they can import into their internal systems. The work still must be reviewed by an attorney - Claude is nowhere near making the kinds of judgments a lawyer would make about this content - but it has increased their throughput from 2-3 documents an hour to 8-10 documents an hour by killing the busy work.
At my company we're also in the process of deploying a system that does exactly that. And what's interesting is that you absolutely do not need a frontier model for that, a small VLM (vision-language model), with optionally a little bit of fine-tuning, gives you the same output quality at a fraction of the latency and cost.
I’m curious how this increased throughput happens.
You’ve accurately stated that AI isn’t as rigorous as a trained attorney. Doesn’t that mean that every single datapoint must be confirmed by a human?
How is that quicker than just using a human to read the content and make the call? Data entry savings?
Yes, remember that these are effectively random PDFs in various different designs and formats, some of them not editable or even OCR'd.
It took a human attorney 20-30 minutes on average to manually copy-paste data from these PDFs into a spreadsheet (while also fixing any errors they found in the document and re-checking for quality).
Now, the AI copies everything into the spreadsheet in a small amount of time, and then the human reviews it. It takes maybe ~5-7 minutes to scroll to the appropriate pages in the document, read the lines vs the spreadsheet, and make corrections. So you've gone from 2-3 items an hour to ~8-10 items an hour.
Maybe you could pay someone to develop an OCR/ML application that could do this. But that project would never be profitable, even with the time savings. At the cost of a couple Claude subscriptions, it makes sense.
Does the human find enough bugs that they stay on guard, or just rubber stamp everything without really looking at it? It’s hard to stay vigilant when stuff looks plausible.
Complete speculation: you could instruct Claude to hide one random mistake in every document.
And Claude should write down the mistake in a sealed envelope, so it doesn't make into the database.
A review that doesn't find the mistake counts counts as invalid.
Don't forget that humans have a not insignificant error rate when copy/pasting or copy/typing data.
I'm doing some public court records processing for bankruptcy cases (interested mostly to seek out corruption in big national cases), and yes, the "variousness" of random PDFs is exactly the issue. Trying to get the cost for a whole case down to a minimum.
Sample is around 300 court dates, shy under 1k files.
At best I'm building a claude skills file.
> Maybe you could pay someone to develop an OCR/ML application that could do this. But that project would never be profitable, even with the time savings. At the cost of a couple Claude subscriptions, it makes sense.
A better use of these Claude subscription would be to develop the app (which it can pretty much do at that point) and you could iterate to make the workflow even more efficient than your current one.
Nobody working there has the requisite experience to do this in a reasonable amount of time. These are not particularly tech-savvy folks, Claude use aside.
Yes. And it might not even be worth it, as the AI agents gets cheaper and cheaper.
Keep in mind that the task is fixed, so as the frontier of AI advances, you can switch to a cheaper trailing edge system and still get the same or even better performance for this task.
Reviewing something takes less time than producing it.
I...don't think that is universally true.
Yes, but empirically in this case, it seems to be true.
Few things are. It also doesn’t require standards. “Stamp this diff” culture is everywhere even before AI. A stamp is literally easier than anything else.
Whether that is useful measurement I suppose depends on the circumstances.
You don’t need a trained attorney to schematize data. The LLMs are used to make the data easier to understand and manipulate.
They'll also hallucinate and change meaning in the process of extraction and "schematization"
Not necessarily. Depends how you use it.
"Write a python script that breaks down this PDF by X feature" would not hallucinate anything in the PDF. Certainly you could trivially double check that all text in the extracted JSON was in the text layer of the PDF.
How much experience do you have with LLMs exactly? It would be consistent with my experience if Claude stuck in a line of python that just emits a JSON literal with no justification, potentially buried in a large program where an untrained person might not notice it. I don't even trust them if the output consists of structured data paired with source images from the PDF, because I've experienced LLMs fabricating the source rectangles to match the output. I only use tools like this by asking for programs, because as you note LLMs are good at that, and the verification process consists of tool calls to legitimate PDF manipulation tools so I have some confidence everything is above board. Even then I only do this for hobbies, not anything that matters.
Lawyer here. I used to trust Claude as hallucinations are near non-existent now. However for large volume tasks such as due diligence exercises, they still happen. We also tried Legora's tabular review, there were also numerous halucinated provisions in our due diligence exercise.
Junior associates hallucinate too...
And when they do, you can train them or fire them, and they learn not to do it.
LLMs change not a whit, and there's no one to take responsibility for the failure (and thus no way to fix it).
As the new variation on the old theme has it, "A computer can never be held accountable, and so very many people are trying to get them make management decisions."
You can’t train people to never make a mistake, particularly when doing highly repetitive work like this. You must build your systems to account for that regardless.
You can scold juniors and they will learn. You can't scold Claude.
Surely the rate of improvement in new LLM models is the equivalent mechanism?
You can scold Claude. Just doesn't make a difference.
Doing similar-ish things with Claude, it's helpful to have something to ground it.
For instance, if you can say:
"Refer to the database schema in x.sql as your source of truth for the database structure we want to import into*. Do not invent data, tables or columns that do not exist. Carefully match all output against this database schema and do not create output that doesn't exist if it does not match the schema, simply skip it."
You will end up with a far better result in my experience.
Gotta treat it like a child.
They do, sometimes. That's why the review still has to happen.
But now it's comparing already filled columns on a spreadsheet, not copy-pasting every single thing from an (often uncopyable) PDF.
I'm a lawyer myself.
I see problems in LLMs doing research, making drafts, etc.
I see 3 kinds of potential users.
1. Non legal background people trying to avoid going to a lawyer. They should definitely NOT use an LLMs.
2. Fresh out of law school graduates.
They think they can do fast work, draft quick replies, do fast case research, enh. They should go the traditional route of interning, learning the "art", they probably won't need it because they will learn "how the law works"
3. Established attorneys. They usually have templates made for most things, over their career, they have established routines of making things fast, maybe it can help them but to say replacing good chunk of work or offloading to LLMs isn't probably going to go well.
There is a reason why interns are made to do grunt work. The "chain of command" is built so they learn on basic stuff, learn to make decisions, etc. Without these basic decisions under their belt, an intern can't take bigger decisions later on. They just won't know....
>Non legal background people trying to avoid going to a lawyer. They should definitely NOT use an LLMs.
I used LLMs to replace a lawyer this year to great effect. I successfully advised myself as to the right strategy and drafted an immigration petition in the correct language. Granted it wasn't 100% LLMs (I also supplemented it with "traditional" online research and wrote the letter myself using the LLM version as rough guidance), but still, I certainly would've gone to a real lawyer in 2024 for this stuff but didn't.
Professional advisory work can be divided into two types:
1. Once you know the rules of the game, the next move is immediately obvious.
2. Knowing the rules is not enough to know the next move -- that requires judgement and experience.
LLMs in the hands of someone smart and savvy can easily replace the first type of work.
If it is at all possible to replace a lawyer with an LLM, it should always be the preferred choice. Using a lawyer is not a zero-sum game, using an LLM is.
Lawyers nearly universally tend to: convince clients they need a lawyer; promote extreme views; charge legal fees (that ultimately result in financial damages to the client). There are no winners in this game, except for lawyers, who are having a lovely time. Even if you won the case, you've lost, because the pie got smaller due to lawyer fees.
Is that based on personal experience, or just anecdata?
> There is a reason why interns are made to do grunt work. The "chain of command" is built so they learn on basic stuff, learn to make decisions, etc. Without these basic decisions under their belt, an intern can't take bigger decisions later on. They just won't know....
That might be true, but that doesn't mean you benefit from training up juniors.
> Non legal background people trying to avoid going to a lawyer. They should definitely NOT use an LLMs.
In my experience, it is incredibly hard, practically impossible, to find a lawyer who will even talk to you. I think this is a valid use case to get at least some understanding what the next steps might be and if it even makes sense to seek legal representation.
Agree. Now imagine the same thing in less developed and/or less well off countries.
There was a post here a few weeks ago about a UK court being inundated with some kind of claims (unfair dismissal?). To submit a claim you do not need a lawyer, but the process is (intentionally?) vague and very complicated, so it takes a lot of effort to figure it out yourself. But AI makes that kind of stuff easy.
AI is not going to replace layers billing big corporate clients millions a month, but it will replace things that are fairly standardized processes, that don't require a lawyer, but today are just too complicated for regular people.
> AI is not going to replace layers billing big corporate clients millions a month, [...]
On the margin, AI will drop these guys' billable hours.
> To submit a claim you do not need a lawyer, but the process is (intentionally?) vague and very complicated, so it takes a lot of effort to figure it out yourself.
I'm not sure there's quite accurate. The form is really not that difficult to complete.
It's basically: Party details, type of claim, particulars, damages
Obviously knowing these details requires some legal knowledge but I actually think ET claims are easy enough to bring if you've got the time to do a little research and build your case. Almost all ET judgments are public so there isn't the usual difficulty with research as you get with law reports being restricted either.
https://assets.publishing.service.gov.uk/media/65bcbd214a666...
> In my experience, it is incredibly hard, practically impossible, to find a lawyer who will even talk to you.
Really? I find them quite talkative. But I'm not sure what kinds of questions you are pestering them with?
In any case, the whole discussion reminds me of people using LLM to replace medical professionals.
For both: unless you have a doctor or a lawyer always on standby, you have to make a decision on whether to even go and seek out one of these professionals. By definition, you make that decisions without professional help. I think LLMs can help you there with the initial research to decide whether it's even worth it to contact the professionals.
Might be true today, but I still feel pareto principle applies. These points also sound like gatekeeping a bit and I hope future AI versions prove that law is also essentially something that can be made as programmable as possible for most common answers if the underlying principles and constitution are organised such.
If there is no precedence for X happening when Y , Z, P, Q, R are special combinations , the judge decides and sets precedence. But if another judge interprets it differently with similar set of circumstances it is not considered incorrect I presume especially if in different jurisdiction unless overridden by a higher court. This is my layman understanding but I think this design is wrong and essentially where human subjectivity controls the system.
Subjectivity is not totally avoidable in law but the goal of the legal system should be reduce it as much as possible along with ambiguity.
Your understanding is correct, but where your understanding goes astray is thinking that that's a bug. The subjectivity is a feature. It is not possible to specify what should legally happen for all possible circumstances the entirety of humankind faces. Like, literally not possible. We humans can't even create rigorous specifications for what our programs should do that keep up with their evolving requirements, and you want to make life-ruining decisions set in stone based on a rigorous set of predetermined rules about all possible circumstances the entirety of humanity could face any time in the future in an evolving world?
Law is intentionally not code. The world is not something you can program.
Isn't achieving a rigorous, unique specificarion what civil law tries to achieve? Not that it is a success, but the idea behind it. Sorry if this is a silly question, but this is how I understand it.
No, it doesn't. It aims to provide a fairly comprehensive framework within which judgments can be made, but it is absolutely expected that there are edge cases and ambiguities which are resolved by human judgment of the circumstances.
Since this is a thread about using ChatGPT for law, I should note that this is actually one of the most important reasons you'd rather hire a lawyer than ChatGPT to represent you. A good lawyer or firm will know the local judges and how they tend to lean on ambiguities in their area of expertise, and will be able to use that information to both advise you and tailor their arguments to appeal to the judge's sensibilities. There is an element of unfairness to this, to be sure, but you'd find much more unfairness really quickly with machine-generated judgments because of how messy and non-compressable real world circumstances are.
I understand, thank you so much for the detailed answer.
Write me the code that says how the Lindsay Clancy case should have come out.
if murder: verdict = guilty
You can replace "laywer" with "software engineer" in your post and it holds equally true. I think this also applies to other factions.
I do think non legal background folks can use LLMs today to sense check legal ideas, like for instance, "what are my legal rights in this situation?"
These situations usually are not ones that an individual can justify the time or money to contact an actual lawyer, but then if they do decide to contact one they will come in with better questions and more of a sense of what they are expecting.
This is similar to medical. Should you use LLM to diagnose yourself, treat yourself with prescription drugs you buy from shady gray market online sellers? No. But you can use it very well to know when it's time to go to the doctor and what to ask.
Yes. Unless you have a lawyer or doctor on standby, you have to make the decision whether to contact them. And by definition you have to make that choice without professional help.
LLMs can help with that. I don't think they are worse at this than me trying to figure this out all by myself.
>I see 3 kinds of potential users.
3 kinds of users that don't scare the shit out of you.
My buddy works with lots of folks that have gone all in on Harvey. There are other kids of users.
There are 3x as many lawyers per capita in the US as we had in the 1970s. Credit agreements that used to be 50 pages are now 500. This increase in volume has not lead to any increase in actual value. The legal profession is a metastatic cancer on our society. Un-fuck this situation and maybe I'll think about taking your advice of not using a LLM.
Another thing that happened in 1971! https://wtfhappenedin1971.com
That's website is a great showcase of how to lie with statistics.
That very first image uses a different inflation indices for each of the two lines in the graph.
They also only post pictures without linking to sources.
Once again, LLMs comes to solve a problem that should not be there in the first place.
They'd be better off outsourcing to lawyers in another country.
My dog can review documents at an even faster rate.
You havent given any proofs or even comments that the work is the same level of quality or accuracy.
I'm thinking back to some of my interactions with corporate lawyers and you know what? My bar (<- pun) for trusting Astra more than a lawyer is pretty low.
Visit a developing or corrupt country and you will realise the LLMs of today analyse better than the judges from purely precedence and literature review POV
The proof would be that the attorney who did the work before, and who still reviews all ingested data, says it is.
Exactly.
The statements in the post are opinions, there is no actual PROOF they are true, and thats my point.
You should free your dog then because a creature that intelligent should not be anyone's pet.
And free the biglaw junior associates while we're at it!
They're lawyers. They don't deserve freedom.
Lawyer here (non practicing so to be clear none of this affects me):
most comments I read here don't seem to realize that different areas of law have very very different economic models and don't even mention which one they think will be affected or why, they just sort of lump it all together.
For example: It is highly unlikely llms will have any meaningful effect on high value personal injury law - I don't see a 5 million dollar case being handed to an LLM when the majority of the cost is in trial aids and not even lawyers. It may affect where and how they advertise. It may affect how they work. But it seems really unlikely to put any of them out of business any time soon by people doing it themselves.
Will it affect other areas more? Maybe. Probably? But so far I haven't seen a ton of comments that make specific enough arguments that they could really be debated or responded to effectively with a useful opinion
> them out of business any time soon
No. It'll be like software. Entry level employment will be affected. You wont want or need associate attorneys when you can hire a brilliant AI associate for 1/10th the price.
> You wont want or need associate attorneys
In commercial law, the business model is to rent the associates out to clients at rates higher than you pay (d/t your supervision). No client will pay $350/hour for your AI.
And for direct billing, it's not your interest to do things faster for the client (at risk of making a mistake).
So the real question is whether these pay-by-the-hour models will die as a result of competition from fixed-price transactions made viable via automation. I think not because the entire difficulty of legal work is that it's a complex bundle of issues until you sort them out; once sorted, there's really little to do other than execute (read: paralegals).
Corporate lawyer at big law firm here. I believe the model will be like this: we will review your 1000 agreements but with AI. AI is generally accurate but has some risks, if you agree we will charge you less. The client will agree. Instead of spending 300-400 hours to review the ageeements AI will do instantly. A lawyer will spend the time on reviewing the red flags identified and draft a report. We already do the same with translation. Before computer translation became good (i.e pre-Deepl era), we used to translate documents (cease and desist letters, court petitions etc.) ourselves. Now we offer AI translation free of charge. If client wants reliance on the translation we charge a small bit. Lawfirms are also under pressure for meeting internal budgets. Most clients ask for capped fees. If you charge 10k for a job to a client and the hourly costs ramp up to 20k, the management criticises the partner, the associates become unwilling to work for the partner (as their hours get written off). So with AI law firms can bet more competitively as the cost (human attorney hours) are reduced.
> No client will pay $350/hour for your AI.
But they may be willing to pay $200/hour for your AI (plus, of course, your final expertise in the matter as the lawyer), and the firm can ditch the paralegals and pocket the difference.
The point is, if you need fewer grunts doing the busywork, you can charge less and be competitive.
There's a middle area too between your options - a law firm that utilizes AI offers cheaper pricing than the current firms for the same work. Ideally competition will drive prices down but of course law doesn't have the benefit of do-overs on cases, so reputation still plays a big factor.
That's the big question, how prices evolve if LLMs increase supply. Standard economic theory says that prices must collapse. There is probably still a short time of 2-5 years where AI native firms can pocket in quite some arbitrage but I guess markets will eventually figure it out and prices will collapse. This is not only for law. The same applies to all the other industries where LLMs can be used to automate work tasks.
> No client will pay $350/hour for your AI.
But they already do! There are cases of professional service firms creating their reports using LLMs, there have even been a few prominent scandals re the fake data generated in them
It is in litigation as well. The issue is that corporate clients are demanding AI. They don’t want to pay associates to do a first draft, they want senior partner to just push the ChatGPT button and then check it.
> No. It'll be like software. Entry level employment will be affected.
Whether everyone at the labs knows it or not, this it what they are doing to entire industries, starting with software: Embrace, extend, and extinguish
https://en.wikipedia.org/wiki/Embrace,_extend,_and_extinguis...
What a time to be alive!!?
It'll be like software. The demand for both software and legal services is effectively infinite. If it becomes cheaper to file lawsuits then we'll get more lawsuits, not the same number of lawsuits at a lower price.
There will still be jobs for competent associate attorneys to prompt LLMs and review output. But frankly a lot of attorneys — especially many of those who attended "Third Tier Toilet" (TTT) law schools — were always a waste of money and should have never entered the profession in the first place.
There is a huge problem in at least family law where a tactic is to simply ddos your opponent with motions and frivolous complaints until someone runs out of money and settles. I'm watching someone finally successfully defend themselves using a Harvey like tool they built themselves. An appellate attorney they consulted with recently, told them they had never seen anyone going pro se get as far as they had in appellate court. So I'm hoping that this serves as a return to making the legal system more accessible, and not just something used to railroad people.
You can’t 10x the number of lawsuits, as the courts are already at capacity. If frivolous lawsuits become common they’ll just change the laws to require a greater damages for it to be heard.
Lawsuits are just one type of legal services. There will also be more contract reviews, more cease-and-desist letters, more patent applications, more regulatory comments, more license applications, etc. Although I do expect legislatures will eventually budget for expanding civil court systems in response to lobbying from business groups to speed up trials.
Can AI replace divorce lawyers? Marriage has no friction, but divorce is high friction. Made impossibly expensive and time consuming because of the cartel of divorce industry.
There are hundreds if not thousands of divorce lawyers. How is it a “cartel?”
Divorce is expensive because divorces are acrimonious and people in that situation would rather spend the assets fucking each other over than letting the other person get them.
In France many divorces are uncontested, but you're still require to use a lawyer (the same for both spouses).
On the other hand most couples with kids are not married, so the separation doesn't require a lawyer...
Divorce is expensive because it is big bucks for the lawyers and family court cartel. They have a vested interest in not changing the laws. When laws are made fair, and the perverse incentives to divorce are lessened, divorce rates plunge. However, divorce lawyers lobby against these laws.
Kentucky became the first state to pass a law making equally shared custody the default arrangement in divorces and separations.
https://www.wsj.com/us-news/law/the-equal-custody-experiment...: Divorce Plunged in Kentucky. Equal Custody for Fathers Is a Big Reason Why.
"Cartel" has a meaning. Explain how either of those things is a cartel?
AI will likely replace divorce lawyers (or, at least make existing divorce lawyers much more efficient) - for non-contested divorces that are amicable. Roughly 95% of what a divorce lawyer does in that context is entirely automatable.
amicable divorces are like 5% of all divorces, so there's not much to be won there
You may have misread that statistic, it's exactly the opposite: around 5% divorces are not resolved amicably and proceed to be litigated in court.
Divorce is extremely cheap if you both agree on everything. Its literally one of the options on the divorce filing. You just pay the court fees. The problem is both parties rarely agree on everything and that's when it gets expensive.
There’s an argument to be made if divorce was easier than more people might get married when they are less serious. My view though is make it harder to get married and easier to get divorced.
"Marriage has no friction" is only something that could be written by someone who was never married.
He clearly meant getting married has no friction, while getting out of a marriage has quite a bit. Presumably there’s friction in the marriage otherwise why would they get divorced.
> Made impossibly expensive
huh? Many/most are no-fault jurisdictions with strict rules on dividing community property. Divorce is easy and cheap, unless it's hard to find assets.
It's not just the divorce lawyers. We have 3x as many lawyers per capita in the US as in the 1970s. The system is designed by lawyers for lawyers. They will create more bureaucracy and more procedure to consume any surplus.
> the cartel of divorce industry
So like a group of lawyers among the "divorce industry" who set prices for divorce and stuff? or wtf do you mean?
Personal injury attorney. AI is entrenched of every single phase of litigation, from the busywork of getting records, research, developing case theory and themes, expert selection, expert collaboration, negotiating, drafting demands, drafting pleadings, drafting discovery, drafting discovery responses, deposition prep, motion practice, jury selection, voir dire, opening statements, witness direct outlines, witness cross outlines, closing statements, and appeals.
Yeah, this matches my friends - they use LLM's for tons of things, using Eve and such. But nobody is replacing the lawyers themselves in cases of any value.
I also have friends in the world of high-volume personal injury (IE the morgan and morgan's of the world) but they also aren't being replaced by LLMs so far.
Emphasis on "so far"
I'd love to get your thoughts on the current frontier models for law, or a link to anything you've written on the subject. Where do you think they're at?
LLMs will never replace human experts.
The word “never” is a dangerous word. In my lifetime I have seen it many times where something was supposed to “never” happen only to see it happen a few years later.
I'll buy that - but, how often does one need an honest-to-goodness expert? 99% of the time most people's requirements can be satisfied with general-competence, and, while the AI systems we have today are mostly only generally competent in porting code from one language to another - they are getting consistently better. I expect we'll start to see general competence in many more areas in the next few years.
> I'll buy that - but, how often does one need an honest-to-goodness expert?
Any time you are doing anything important, where the outcome actually matters. For example, I was recently applying for a study permit and if I generated my application using an LLM instead of consulting with an immigration lawyer and writing it myself, I would not have gotten my study permit. I was recently applying for internships, and if I hadn't paid a real recruiter to review my resume and advise me, I would not have gotten my interview. And trust me, I spent at least an hour trying to get an LLM to generate what I wanted, but no matter how I begged or described what I wanted, it was no use. Good luck fine tuning an LLM to be good at any of this.
> they are getting consistently better
No, they are not getting better, and they will never replace an actual human expert, because a human expert has the actual experience and up-to-date knowledge of what they are an expert at, and also all the complementary context and the current state of affairs in that area. The LLM has none of that.
LLMs are replacing human experts
in cases where the quality of the output doesn't matter
Many people have very low quality thresholds
Then hopefully at some point they get to experience the consequences of having low quality thresholds, and perhaps then their quality thresholds will increase.
This is a weirdly reductive take.
There are plenty of areas where a "low bar" is more than enough. "low" doesn't necessarily mean bad here, it can simply mean "not the best".
I would argue the majority (in the true sense of the word) of white-collar service work does not require an expert.
It requires someone who has reasonable qualifications in the given field, and the motivation ($) to help you with your task.
So no one needs a top-shelf $900/hour lawyer to draw up a basic low-stakes contract where both parties agree to all the terms as written. You just need someone to ensure that the contract is legally valid as written.
Are implying that an LLM can accurately evaluate whether a contract is legally valid?
Or in the long term, no new experts are being made because no companies want to sponsor the slow work and resources that becoming an expert take
Are you aware of what frontier AI has done to mathematics in the last 6 months?
Lol no one cares. Most of that is smoke and mirrors anyways. Stokes was plagiarized. Most of these "breakthrough" solutions are actually present in the training data and in the prompts researchers and mathematicians input into ChatGPT themselves, prompts OpenAI then steals even when you don't agree to improve the model.
PS: I use LLMs every day, basically let Codex write most of my code (after 15+ years of professional programming), these are amazing tools, smart in some ways, but completely demented in others. People just need to fucking stop being AI doomers and AI coomers. Is this technology great? Yes. Is it worth 10 trillion? Hell fucking no. If we get to a point where RSI is a thing, then maybe. So far it's not a thing and its not even close.
I suppose not. I am not a mathematician, and there has been too much drama to understand anything.
There's a huge problem on here - many people stepping outside of their domains of expertise with surface level knowledge.
Its teh same reason claude for finance hasn't turned the finance world upside down.
This is getting tiresome seriously. Why wont these geeks learn some lessons?
Yeah geeks just stop trying! Seriously? Yeah, we’re in a bubble and narratives are ahead of reality but if you really don’t think AI is and will continue to eat knowledge work, just keep making your buggy whips.
Many people stepping in their domains of expertise with surface level knowledge
i agree with this, maybe.
One of the things I've been noticing are lawyers getting annoyed when they realized you spent time working a matter with an LLM. In the cases I've seen, it had nothing to do with the quality of what came out of the LLM. It seems to be related to --and this is only my perspective-- a reality where now reasonably intelligent/experienced clients are coming to them with a case fairly well digested, whereas before they would have to pay the law firm tens of thousands of dollars to ingest documentation and information and do the digestion.
In a couple of instances I've come to the table better informed about the relevant law and case history than the attorneys were. In many ways I think this is a matter of people understanding that none of us are infinite human databases, so you can't know everything. Instead of rejecting LLMs (perhaps being passive-aggressive), they should embrace them as well as better-informed clients.
Having used LLMs for software development I know full well just how terrible they can be if you don't know what you are doing, don't guide the work correctly, etc. So I am not taking legal LLM analysis as ground truth, but it isn't garbage either.
The need for actual lawyers will persist I think from my own experience. I attempted drafting a contract with some points myself using AI, but after several edits I wasn't sure if it was correct. Sending it to an actual lawyer ended up in so many corrections I couldn't imagine the first time. One big thing was the overly excessive protective clauses which didn't make sense for reality or conflicted with another.
Its just like code I suppose, if you can read and understand and validate, you can use it to scale and otherwise it could end up being a vibe effort.
I'd correct this to "The need for actual GOOD lawyers will persist I think from my own experience."
I have worked with many lawyers in my career. The bad ones will look at a contract and tell you fifty things that can be improved. Meanwhile, you take the same contract to a really great lawyer and they tell you that's it's not worth fighting over the changes and to just keep it as is. It's kind of like how a beginner or intermediate software engineer might tell you how you can build out a k8s or auto-scaling fargate cluster for your project and a really senior engineer might look at the same thing and explain that it's not worth worrying about scalability right now.
I find that when I feed contracts to ChatGPT (which I do all the time) it will try to find 50 things wrong with it. For example, as a test, i fed chatgpt an employment offer from a US government agency. It came back with 1152 words and 14 recommended (some of them substantial) changes. Meanwhile, any half-decent employment attorney would probably laugh at me if I asked them to redline an employment offer from the State Department.
I’ve found that good lawyers are about getting things done. You could nitpick the terms of a contract, but would you rather buy the house or have a perfect contract? Once I realized that, my view of lawyers changed (for the better).
YAGNI isn't just for software.
In my recent experience, it seems like LLMs may even increase need/demand for actual lawyers (more ideas + more words = more lawyer time). I expect lawyers will generally adopt the technologies that will benefit them, rather than the technologies that would increase standardization and trust
The same has happened with modern finance. Despite more technology that has power to drive more transparent/efficient markets, we have less perfect information sharing and a larger group of middlemen capturing a growing share of profits of the economy
I want to take a step back and observe that you're saying expensive specialists will continue to be required if two parties want to make an agreement. That seems like something worth fixing rather than just accepting.
> I want to take a step back and observe that you're saying expensive specialists will continue to be required if two parties want to make an agreement.
A lot of contract-lawyering involves getting ahead of when things go south and the two previously agreeing parties find themselves in disagreement.
Same for SWE. AIs will churn out happy path vibe slop and fail when something unexpected happens by swallowing errors or having a bunch of terrible "fallbacks".
The problem is those two people are in conflict not collaboration.
If the expensive person is still better at getting what _you_ want from a contract then it’s always worth doing that.
There’s a reason lawyers are so expensive to begin with - they do very complicated stuff with very expensive consequences if done badly.
It's probably somewhat of a zero sum game. Both sides lawyers have AI tools which probably just keeps them on a level playing field rather than actually making anything cheaper or faster.
I spoke to a very very high up corporate lawyer specialising in tech - so knows Ai is real. And they said AI contracts are the bane of their existence. They’d rather you arrive with no idea what you want than a contract partially written by AI. They said it usually takes longer with an AI starting point than starting from scratch.
Millions if not billions of agreements are made every day without the involvement of lawyers. Lawyers are only needed for particularly complex situations and even there they are generally only "needed" in the sense that parties feel they would add value.
Yeah, and you can already get fill in the blank legal forms for a lot of basic situations, including leases, wills, employment agreements, etc.
The interesting question is whether AI can fill a niche these can’t — or, alternatively, help a layperson evaluate a contract they’re asked to sign better than they could themselves.
In my experience, those expensive specialists help me protect against things me and the counterparty cannot even imagine when making an agreement. Do you mean that this would be fixed by AI or by making simpler agreements?
It can imagine what yourself report that you cannot :)
How?
Contract law has evolved because transfering the ownership of things is fraught with difficulty. Ownership as a concept itself is notoriously hard to define.
Consider housing - when you buy a house you own that land. Except if its an apartment, you just own the internal walls. Or when the government aquires your land through eminent domain. Or mineral rights - which don't belong to you, you just own the top 5 feet of topsoil. Or if you neglect a property and some squats in it. Or if you were given a fraudulent payment, and they take back the property.
Just getting people to agree is notoriously difficult when people have different often conflicting desires.
Most people dont appreciate that the world is a complex place. Nothing is ever simple, sure concepts can be extremely simplified but the nuances and details are lost. These details are the important parts, just ask a surgeon or rocket scientist etc.
LLMs are the first genuinely useful legal tech since the Internet. I'm pretty shocked, though, at the delta between how competent Claude is on code versus legal work. It's good for research and data organization, but terrible for drafting. I wonder if this is a structural problem with the lack of feedback loops. In law, there's no compiler to check for logical or continuity errors in your brief, and there's no unit tests to check for correctness or performance.
Even without that, I think it'll be extremely valuable to clients to allow them to answer simple questions without a lawyer, figure out the lay of the land so they can supervise their counsel, etc.
You should know - for coding they make terrible mistakes as well.
But programmers have this concept of a "code review" where another person looks at the code to look for problems.
We use this same technique with our LLMs. Most mistakes are caught by having a second LLM look at it. Doesn't even need to be a different model - just make sure it has a different context.
> Doesn't even need to be a different model - just make sure it has a different context.
I find that the number of issues discovered is noticeably higher if you do use different models though. I'm doing some very finicky things (formal semantics) and find that there's value in review panels as large as 5 different models from different families. It gets even more profitable if you set it up as a truly agentic panel where after writing their own separate reviews they get to see what others have said, and adjust their positions or defend them etc. Some models are not that good by themselves but can be surprisingly good at finding flaws in others' reasoning (Grok for example).
Programming has a long standing culture of accepting the code to be somewhat wrong, so we have various tests, linters, reviews and error handling. Also in programming there are many ways to do something and it's the end result that matters most.
Not so in other knowledge work. There's no test harness for a contract and error is non-recoverable. Likewise in finance. There are specific ways of doing things and these ways are many times set in regulations. LLMs can assist all day, sure. But replacing the human, in highly regulated, zero tolerance for error environment?
You could say the Salvatorian Clause in contracts is like exception handling: a "catch (all)": even if some clauses in this contract are illegal, the remaining contract stays in place.
Logically, this actually doesn't make sense strictly speaking because the sentence creates a paradox: doesn't it make clear whether it includes itself or not, and each reading ends up in trouble. There is a "tradition" in law around the world to accept the only benign reading of such clauses, which I always found funny given that in all other ways lawyers adopt the most adversarial mindset imaginable.
I don't understand. I just provided advice about getting better output. Are you trying to reply to someone else?
You made some points worthy of expansion:
>> You should know - for coding they make terrible mistakes as well.
>> But programmers have this concept of a "code review" where another person looks at the code to look for problems.
> LLMs are the first genuinely useful legal tech since the Internet
That is an incredible statement that could not be further from the truth. Large scale adoption of email, searchable document databases like Westlaw, LexisNexis, PACER, etc.. , OCR Software, electronic signatures, and tons more have had a much more defineably positive impact on the legal profession since the internet came about.
I think "email" and "westlaw" fairly count as "the Internet." LLMs might be bigger than either of those.
The internet came well before both of those. And none of the example I gave have lead to lawyers regularly being fined in court because of "hallucinations"
I have found that it’s useful generally speaking to get the intent of contracts and red lines, but actual drafting I agree is where I lose all confidence. My guess is that the significance of the difference between using a word like “and“ or “or“ can be so meaningful that that level of nuance can often be lost. But I know nothing I’m not in the space, I just pay too much money for lawyers.
Law is certainly a less verifiable domain
Not trying to be dismissive here but a lawyer will always edit your proposal. I do think lawyers will persist but not as many. And they will work very differently, much like we already use claude / codex to code - a big part of the contract probably won't be read by the lawyer.
I would rather argue they SPECIFICALLY will read everything. However, they'll likely often just be like "I'd phrase this differently but that works, too".
Have you watched Nathan for you? There is a lawyer that didn’t read on tv. And almost got in trouble.
Human professionals put their reputation and finances at risk when performing their work. This risk functions as a guarantee.
> Human professionals put their reputation and finances at risk when performing their work. This risk functions as a guarantee.
Do you know how difficult and costly it is to sue an attorney for malpractice?
Do they?
Not being snarky here but I've encountered some really crappy lawyers and they seem to be doing great ($$$).
It really sucks one cannot publish something like "this guy charged me $5k, wrote my name wrong and forgot important paperwork the day we showed up to court" as that could make you liable for damages. There should be a safe harbor for that.
Assuming you are in the US, there is a safe harbour called "opinion based on disclosed facts" for that. You could also fairly argue that the conduct of a lawyer is a matter of public concern
> It really sucks one cannot publish something like "this guy charged me $5k, wrote my name wrong and forgot important paperwork the day we showed up to court" as that could make you liable for damages.
That's pretty much what bar associations are for, filing a complaint against an attorney at the bar seems like a pretty standard procedure.
>There should be a safe harbor for that.
A safe harbor would be the opposite, it would be immunity for the lawyer.
Even if there were it would just be another court case
Also, unlike SW engineers, lawyers control their industry and will quickly circle the wagons to protect themselves. Paralegals are probably screwed, but lawyers will do fine. I'm considering law as my post-ai-SWE-apocolypse career path.
I’m thinking asking the same lines, but also looking into research / lab chemist so I can hang out with my best friend. If im going to change careers so drastically, I might as well enjoy the company.
I was in a similar situation and tried a different approach.
I had start with asking for a contact with some details I provided. But then I ask the model to be an experienced corporate lawyer and ask me a series of questions to gather the details it needed and then write the contract. The result covered a lot of details that were highly relevant but were absent from the original attempt.
The key insight is that you can lean on the model to cover your unknown unknowns.
Your comment is a great example of the phenomenon where people think AI is an expert in areas that are not because they have no way to determine if the seemingly good looking result is genius, gibberish, or somewhere in between.
The point was that you can go further than "write me a contact that says x,y and z". I actually believe the process I went through with the AI was a more custom, tailored experience than any interaction I have had with actual lawyers, who'd usually apply a template. The AI asked me a bunch of questions, many of which were relevant, and led to follow up questions. I could ask my own questions, make corrections, etc. and all along, they made sense to me. This part of the experience was great.
There could still be more unknowns that I did not encounter and by definition don't know about.
And the final output could be wrong, wording, exact legal terminology etc.
Can we say working with an actual lawyer is guaranteed to be better, more correct? Lawyers are people and they also make mistakes, wake up on the wrong side of the bed, hate their job etc, just like the rest of us.
Here's another techniqe that gets you (even) more out of LLMs: run each prompt several times and compare the various outputs. It is not uncommon for models to contradict their own advice, and also you will get additional insights not included in previous runs.
This is due to the fact that LLMs are statistical processes that rely on pseudo random numbers in chosing what to say and how to say it to a substantial degree.
> The result covered a lot of details that were highly relevant but were absent from the original attempt
Another one suckered by the plausibility engines.
Did you validate the output with a lawyer?
No admittedly not. I can try it and report back.
> The need for actual lawyers will persist..
I agree. But, we will need a lot fewer of them. My small company called our lawyer a lot more before the advent of AI. Mostly to clarify a few things or review simpler contracts. Now, for simple things, we do not call them at all.
Their work will get cheaper so it will become more accessible for people, increasing demand over all would be my guess to some degree at least.
I anticipate an effect on society that troubles me:
(1) Rich people will use human expertise and "the rest of us" will use AI models to get by. This could happen in law, but also in medicine (in particular in societies that - like in the US - do not have universal healthcare).
(2) A positive effect in both domains may be that access to knowledge will be broader and cheaper.
I hope I will be wrong about (1) - I would not like to live in such a two-class society.
you can't in imagine the same way you couldn't imagine generating the text from your initial prompt. when will the hubris end?
some variant of "oh i am capable of generating text i know nothing about, but it helps to have a human i can rely on to tell me more about the text i know nothing about."
pick up a bachelors degree worth of books to humble yourself
> One big thing was the overly excessive protective clauses which didn't make sense for reality or conflicted with another.
I'm currently negotiating with someone who I feel is clearly using AI and this is EXACTLY what has been happening
> The need for actual lawyers will persist I think
But will their glamorous salaries persist? That is the question that matters.
AI doesn't need to wipe out lawyers. If they just depress salaries enough, virtually nobody is going to want to be a lawyer anymore.
The salaries are a result of the insane hours they have to do. Biglaw needs to pay those salaries to keep getting young associates in the door because burnout is a real problem in the industry. If salaries were lower and hours/stresses were lower, you would probably still find people wanting to be lawyers. A different kind of person, maybe.
(Not that I think AI will necessarily have that effect. We just don't know yet)
Salaries are about scarcity and leverage, they have little to do with working hours. Nurses pull insane hours and they still don't get paid as much as doctors.
Very few people, if any, are going to bust ass in law school or medical school if they're not going to get rich off it. Better to do literally anything else.
It’s a professional organization that only allows so many people to pass the bar per year so I don’t see that changing anytime soon… if anything if they need fewer lawyers they will just pass fewer people to keep their wages high
> The need for actual lawyers will persist I think from my own experience
This doesn’t replace lawyers, but paralegals surely will be affected. A good enough model could shrink the number of paralegals needed in a firm.
As costs for legal work drop we might expect demand for that work to increase. For example, it may be the case that legal help becomes accessible to entire segments of the population that had no access previously.
The best writer I know of was an associate attorney. He didn't have the technical background nor the in depth computer related knowledge, and relying on the information I fed him. But man the briefs he filed to the court were amazingly good. Reading them I would have been convinced his side was right if I were in the jury.
Law LLM will surely help competent lawyers in their fields with greater sources of knowledge not in their core area of expertise.
I don't want to imagine a world where juries were responsible for reading briefs and making decisions.
Sure as long as they stop making things up.
And to add to your last point yes it should also be like code: where the costs of development are coming down. Legal costs should trend down if law firms are effectively leveraging LLMs to accellerate their procedures (though I won't hold my breath)
I got the exact opposite. A contract from a lawyer passed through ai found so many flaws.
> The need for actual lawyers will persist I think from my own experience.
The outcome of a case shouldn't depend on someone's fallible ability to recall facts or convince other people or point their index finger*
Law should generally be deterministic. One's CHA stat should have no bearing on justice.
There should still be human judges, but the middleman between the judge and petitioner could easily be removed, and have generally been seen as leeches since forever anyway.
Though, like how the USA opts to remain in the Stone Age with regard to tax filing because of lobbying by tax software companies, this faction of society will flail the hardest before they admit they're obsolete.
----
* What's a lawyer's favorite programming language? Objection C.
The ability to clearly and effectively communicate, and the ability to know and recall the facts of a case is always going to be important. It'd probably be a more fair system if judges and jurors couldn't see the attorneys or even the people they're representing. Now that many cases are handled over video calls it'd be pretty easy to make that change.
A lot of communication is conveyed through more than the words themselves though and even though it's often misinterpreted people put a lot of value in that information. I suspect that even if we took away the ability to see the people involved judges and jurors would still find bias in the tone/quality of people's voices. Even if everything was reduced to text they'd find bias based on word choices, phrasing, spelling, and grammar.
Yep, just yesterday, Nike removed one leach middlemen called retailers and sell directly through their app. Turns out working out very well for them too.
If you have been following this story for the past couple of years, this has been disastrous for Nike.
The law is much more interpretation and argument based than logic routing based than you seem to believe.
Lawyers sure. Paralegals and assistants are cooked in the way typist and secretaries were back in the day.
Nope. Paralegals and assistants do the type of thing GP hired a lawyer for (and lawyers do, as well).
Lawyers can practice, can represent you in court, etc. that's not going away soon.
Lawyers have that lovely regulatory wall, for better or worse.
The typical pattern is called “deskilling”. It doesn’t usually mean a skilled profession will disappear overnight. Instead, the job might be done by less expensive folks like paralegals.
An example is in the banking industry, where making a loan used to require deep analysis of a person’s credit worthiness. Now they use an algorithm (credit scores) which means someone with less experience can do it.
If law follows the same pattern, a job done by someone making $500/hour might be done by someone making $50/hour.
If you think paralegals + AI can do the job of good lawyers you don't understand the profession at all.
i don't think that's the point anybody is making. "good lawyers" is not the profession at risk here. a top tier courtroom lawyer won't be replaced by AI.
but how much of the revenue of the average law firm comes from that, vs the day-to-day "we need this relatively routine contract reviewed" sort of work?
Yes. And it would be the continuation of a trend that’s already been going on for two decades. When I started out as an associate at a fancy Wall Street firm, we had 30 people working on reviewing documents for this matter even that level of firm today would not be able to justify doing that anymore. We’re going to move from a pyramid model to more of a diamond shaped model with many fewer associated attorneys doing the legal research and information marshaling.
> Instead, the job might be done by less expensive folks like paralegals.
So who is going to define your strategy, represent you in court, adapt the strategy to changing circumstances, negotiate with counter-parties on a mutually accepted settlement? A paralegal? An AI model? Please
Very little of corporate law is in court. The mass of things is “does this contract look good?” And “what do we need to change in our business when regulation X comes into force?”
Yes hence why only 25% of the situations I presented were in court.
Former attorney here. A good paralegal is worth 3-4 junior attorneys. A good paralegal with AI should probably still have a senior attorney oversee them but this will absolutely hollow out firms.
"law" and those who pratice "law" have managed to place itself largly in a monopoly. This is a huge drain on the economy. Law is expensive and irrational.
He just gad better prompt
What model did you use?
It's worth noting that the announcement seems to aim the product at lawyers or students rather than laymen.
Our entire world is about reducing the friction of conducting a transaction
How to fix the lack of trust which is needed to conduct a transaction
Contracts affix a moment of trust in time, so even when trust is lost between parties the contract has guardrails for behavior and consequences for acting outside those rails
There are many transactions that are completely inaccessible to most people because the need for a contract or lawyer is too high
LLM use allows for those inaccessible transactions to decrease, and get people further along to the point where a lawyer is accessible as well, and economically viable to use
> One big thing was the overly excessive protective clauses which didn't make sense for reality or conflicted with another.
>Its just like code I suppose, if you can read and understand and validate
I share this, when I asked an attorney for a contract, they were much simpler when compared to self redacted or AI redacted contract, it was as if I were rediscovering lots of concepts and I HAD to put them in paper, even if they were defaults and were essentially no-ops.
Comparable to:
> Mkdir $PWD/path
instead of
> mkdir path
or conditions so rare that they don't merit including.
> x = 0
> if x!=0:
> raise Exception("Bit flip by solar rays or otherwise")
>The need for actual lawyers will persist I think from my own experience
At the very least because the attorney monopoly assigns a 5ish year of training as PoW and natural personhood as Identity as protection for Sybil Attacks.
If anything, the exclusive right of attorneys to represent clients in court, file motions, and enjoy professional secret, is more valuable, as it's not something AI can ever compete at. (Barring a wild recognition of machines as humans, or an overturn of the impossibility of companies to appear in court.)
Second paragraph:
> API customers including Harvey and Legora will be able to build on Astra for Law, bringing this intelligence into their own products and workflows.
In other words: "no, no, we're not eating our children to prep for the IPO. Don't worry."
Max Junestrand has consistently said Legora treats the model layer as swappable, selecting across frontier providers rather than building the product around one.
OpenAI didn't need to name Legora and Harvey in the second paragraph of the launch post.
They are pre-empting the obvious interpretation of Astra for Law: that moving this far up the legal stack puts them in direct competition with their biggest legal AI customers.
“Don't worry, they can build on us” is a pretty conspicuous message to include on launch day.
They have clearly thought about some pessimistic outcomes.
Switching from ChatGPT Enterprise to Legora at my firm was a godsend, it's so much better for legal work, even with the frequent changes to the underlying models.
>Max Junestrand has consistently said Legora treats the model layer as swappable, selecting across frontier providers rather than building the product around one.
Vendor-neutrality for LLMs is such a weak thesis all around, whether for providers or consumers. It weakens the product by being promiscuous and gains no material benefit at all.
LLMs are magic byte(byte) functions, it doesn't make sense to say "we have different providers for magic".
Different flavors of magic require different ritual components and in the name of all that is stable and production worthy, please don't get a necromancer to do your civil construction magic.
Vendor-neutrality helps reduce lock-in, and OpenAI and Anthropic are big enough that the reduction is valuable.
Really? Even when the magic is differently priced across magicians?
"We are also expanding our work on privacy and governance to give law firms specific controls for confidential client work"
This is everything OpenAI have to say about privacy in this announcement. No guarantees. No promises. Just a pinky-swear promise.
Anyone trusting them–or a lawyer who relies on them–for legal work deserves what they get.
the thing is a lot of the legal work which will go through this is drafting 100 and 1 variations of draft versions of standard contracts not containing any trade secrets where the contract can be drafted with "replacement/place holder names"
the kind of work mostly done by juniors not yet through their final exam and other "non" lawyers etc.
so it's a slippery slope of "lets just use it for <this> things where it doesn't matter" and then out of laziness and convenience it creeps into all the other places (at least for drafts).
Lawyers are already llm’ing their work and getting slapped around in court for making shit up.
There’s another case making headlines every week.
I get the feeling a lot of them won’t care about this stuff.
they could potentially lose their license, I think that's worth caring about.
If there is any cartel that deserves to be broken its legal search. May they die a miserable death.
Yeah, it should be freely available, you have to be able to know the rules you're supposed to obey in order to obey them well. I've been making a free API for US law search, you can point whatever model you want at it: https://law.agentlookups.ai/
Very much a work in progress, only federal and state so far, no municipal codes yet, and no case law yet. Big hole, I know. Also working on making the search ranking work better.
is there a dataset/torrent with all the underlying laws?
Right now it's just a bunch of crawlers for the individual states. If there's interest, I could periodically stand up snapshot torrents or something. That something you'd be interested in?
Alternatively, if someone else knows an all-in-one option that exists, I wouldn't mind retiring those crawlers...
Not op, but that's a very interesting proposition. While the law and legal code are technically property of the people, I'm not aware of any single point of download for it all.
There’s no single point of download for it all because there’s thousands of autonomous entities that issue law and adjudicate cases, at least 51 of them distinct sovereign entities.
We could enforce (suggest?) a common format / api at the federal level. Especially if it’s incentivized with funding that more than justifies the cost of maintenance. Similar to how federal interstate funding is only available to states with a 21+ drinking age.
Govinfo.gov is neat because you see the different attempts at common formats that have been attempted in the past.
Yeah, going to the courts and municipal code seems like it's going to be a heavy lift. Many of them seem to hang off of municode, though, so maybe it's not a huge number of unique crawlers.
OK, I just bundled it into a torrent, available here: https://law.agentlookups.ai/downloads/
If you decide to give it a download, let me know if you have any issues with it/suggestions for improvement.
There is none, not for statutes and definitely not for case law; even at the appellate level where you have multiple federal circuits, then 50 states, then territories, military, tribal and a whole host of other niche courts. And the appellate court systems can be split into districts, and by lower and higher levels.
Then if you want to really get into it, The People should also be able to access trial court level, and at that point you have over 3000 distinct court systems with their own access systems, usually requiring logins and CAPTCHAs, and half of them not even having anything accessible online at all, and the other half only having recent stuff online and the rest rotting in a flooded basement.
A third-party project converted an official XML source of German laws to Markdown documents, but the repo is a bit dormant:
- https://okfn.de/en/projekte/bundesgit
- https://github.com/bundestag/gesetze
https://axiom.org is working on this but very early.
Cool, thanks for the pointer!
> By using the legal search index, Astra for Law can search U.S. case law, statutes, regulations, court rules, and administrative decisions across a corpus of more than 230 million URLs, with sources added daily. Our work with Free Law Project, the nonprofit behind CourtListener, brings its case-law collection covering more than 99.9% of published U.S. precedential case law (opens in a new window) into this research experience.
Found that to be very interesting
CourtListener already has an MCP interface and Grok is quite good at pulling from it. In my experience, Grok 4.6 is quite good at analyzing legal cases and human-written documents. Better than Opus 5. I'm not sure if it's better than Fable 5.1 on that task, b/c I'm not willing to spend my precious Fable tokens on case law searches lol.
Let's swap it out for the AI cartel!
The courts are about to overrun with AI-generated lawsuits (even more so than they have been[0]).
[0]https://www.technologyreview.com/2026/06/04/1138391/courts-c...
I find the idea that people can use LLMs to exercise their rights as citizens appealing however. Many people aren't aware of the rights they have, and LLMs are pretty good at surfacing some stuff without having to pay lawyers. Having to hire a lawyer is imo actually a huge way of gatekeeping people from exercising their rights. I heard a lot of local German public institutions are currently being flooded with people arguing their case with the help of LLM that they previously weren't really realistically able to do. So I don't see it all as bad.
>are pretty good at surfacing some stuff without having to pay lawyers
They are also really good at making stuff up as evidenced by the many, many, many examples you read in the news about actual lawyers using AI to write briefs that are full of errors and hallucinations.
In civil courts, you'd likely get more sympathy from a judge if you represented yourself and admitted your lack of understanding, rather than try to appear as someone you're not because you wrote some prompts and copied the output.
Current court systems around the world are just not built to handle the flooding of cases from the citizens.
The stupidest analogy is open source projects having a hard time accepting LLM generated PRs from the masses, because review process is the bottleneck.
No idea how to fix this, to be honest. In coding world, with some mental gymnastics, I can see code not being reviewed by people anymore. In courts, things generally have more consequences, and you can’t really roll back decisions that easily.
you could probably write a cool little gotcha of an SF short story about a barren wasteland of a planet that keeps broadcasting out legalese that's revealed to just be LLM chatbot lawyers pedantically arguing with one another about xeno legal doctrine
Relevant:
https://www.ftrain.com/nanolaw
JG Ballard could probably turn that into something darkly atmospheric, and even more dystopic than his usual depopulated worlds...
The legal system, which is a machine/technology by itself, will be eaten out. I wonder what will replace it. Botnet law arbitrage? – Personal assistants constantly negotiating with each other to avoid permanent civil lawsuits?
the cost of making a legal argument can collapse while the cost of reaching an enforceable, legitimate decision may go higher, which will gate the "justice" system even more.
The legal system is slow and inefficient by design.
It's really not. The legal system is slow and inefficient because it's a deeply pipelined system built to maximize the throughput of the bottleneck resource: judges. Judges are constitutional officers who exercise independent authority in meat space and thus are necessarily limited in number. The rest of the design flows from that.
If you got the judge, all the parties, and all the witnesses in a conference room together until the case was resolved, you could probably handle a lawsuit in a few months. But each judge has hundreds of cases pending before them, so that would never work. Instead, you get something like how a GPU works. You do some work on a case, submit the work to the court, then work on something else for a few months while you wait around to get the results back. Then you do some more work and submit it to the court, then go do something else for a few months while you wait to get the results back. A few months of actual work gets spread out over a few years that way.
Why stop there? The bottleneck is not some immutable force of nature. The number of judges is determined by legislative action. It is within Congress's power to allocate new federal judicial seats, and likewise at the state level with the equivalent lawmaking body (for most states, at least). Why don't they do so?
Because federal district court judges have a tremendous amount of power. You both don't want 10,000 of them running around and you don't want to water down the qualification for the position--if you did do that, then that would devalue their judgments, everyone would appeal, and you'd just shift the bottleneck up to the appellate courts.
At the end of the day, litigation is conflict resolution. You don't just need a decision, you need a decision from someone authoritative enough to bully Fortune 500 CEOs into submission to accept the judgment.
The legal system is robust and deliberate by design. That winds up being necessarily slow. But it doesn't need to be inefficient.
You presented the concern from my adjacent comment perfectly (“LLM performance: common law vs civil law” essentially). So is AI possibly just growing the “Reverence for Professional Experience” factor that plays such a big role for legal compensation here in the US?
I am not sure "eaten out" was the correct phrase to use here.
Call it a Prediction Market
They seem unaccountable. Every other profession has increased their productivity.
justice doesn't have to be productive it has to be fair
It's completely failing at that too. Public trust in the courts is lower than it's ever been, and it was never all that high to begin with
You've never heard the phrase "justice delayed is justice denied"?
how so? delayed justice is denied justice
construction productivity is actually down since the 60s!
>Every other profession has increased their productivity.
You dropped this /s
Don’t you have fee when initiating a lawsuit ?
I don’t know in the USA but in France, if it’s deemed that you launched a lawsuit knowing very well it wouldn’t succeed, you are susceptible to get a 10k€ fine. Even jail in serious cases.
Yes. And you(r lawyer) can collect lawyer's fees and you can be made to pay the court fees, if you lose.
IME the American legal system is set-up to discourage litigation, though. A common tactic is to bury your opponent in the threat of heavy damages or jail-time to get them to settle for what you were originally after, which courts are perfectly happy to facilitate because it gets a potentially lengthy trial off their dockets. They'll punish (or be biased against) whichever party seems responsible for not accepting a "reasonable" settlement.
Algorithmic abuse of the system to extract payments already exists in the form of the debt collection industry.
The solution is obvious: AI judges! With complete immunity for model vendors of course, just in case some toddler is sentenced to death or something.
I wonder how many parties would prefer an AI judge over delayed justice.
Why the hell are they not fighting fire with fire? This is not sustainable. But it is annoying that the AI labs get to play arms dealer, selling to both sides.
Judges don't care if people have to wait months. What does it cost them?
How are we fighting it online? You cannot
It'll be like radar detector detector detectors, which were a thing for a while. Each side (cops, speeder) buying detectors to detect the detectors.
There's a Judas Priest joke in here somewhere.
https://www.damiencharlotin.com/hallucinations/
No word on model hallucinations in the blog post.
https://artificialanalysis.ai/models/gpt-6-astra?omniscience...
Not like common law doesnt already have a lot of hallucination going on.
can “penumbras and emanations” compete with hallucinations?
Common law is all about rummaging around in dead mens’ letters, LLMs are a natural fit.
Genuinely wondering (aka not snarky): Has anyone found frontier models to provide useful research in the context of European civil law systems?
Your comment made me wonder if there are any halfway-acceptable model benchmarks for law tasks? Specifically I’d love to know how the frontier models’ abilities compare between common law vs. civil law systems. My guess would be that an AI in a common law context should have a clearer idea of how a specific case is interpreted/accepted by (common law) practitioners, whereas trying to rely on AI in a civil law context, like Germany, can be daunting. In a few Germany-specific recent examples, the models feel like they present only (maybe too stubbornly?) the “civil law”-based laws. All while negating much of AI’s research benefits because civil statutes are portrayed as being absolutely accurate, binding, and their enforcement (and thereby the legal reality) being uniformly applied. Am I making this interpretation up? If so, how can I prove myself wrong?
I don't know of anything that is working yet, but I know some people working on it (intentionally vague).
Roman law, on which the legal systems of Germany, most of Europe, Turkey etc. are based relies more heavily on statutes than Anglo-American case law, but cases do play a role there, too. That's why a practically useful system also needs to have access to court decisions, commentary etc. - and while the statues are in the public domain, a lot of the other knowledge is owned by specialist publishers (such as: C.H. Beck, Nomos, Mohr Siebeck, De Gruyter, Otto Schmidt, C.F. Müller, Manz, Verlag Österreich, Schulthess, Stämpfli, Dalloz, LexisNexis, Lefebvre Dalloz, Giuffrè Francis Lefebvre, CEDAM, Giappichelli, Il Mulino, Aranzadi, Tirant lo Blanch, Marcial Pons, Dykinson, Tecnos, Lefebvre, Almedina, Gestlegal, AAFDL, Kluwer, Boom juridisch, Wolters Kluwer, Larcier-Intersentia, Anthemis, Sakkoulas, Nomiki Bibliothiki, Universul Juridic, Hamangiu).
Internationally, legal systems appear to undergo a sort of convergence, which means statutes increasingly matter in the U.S. and cases increasingly matter in European law (according to a law partner friend).
The benchmark you linked to shows GPT-6 Astra having the lowest hallucination rate of all tested models.
The blog post makes no mention. If the rate isn't 0% it should be mentioned for a field like this.
Agreed. Hallucinations and reliability are the main hurdles to anything being 'agi' in my book
The definition of AGI is whatever a frontier lab wrote a blog about doing in the last week.
why this maximalism? There's nothing that has 0% hallucination including humans. lets use reasonable baselines.
Humans can be sanctioned and fined, and eventually disbarred if they continue to lie in court filings.
presumably that will remain true for the lawyers using the product
So those lawyers will likely want to know about the likeliness of it, strategies to work around it, and overall mitigations.
It might be helpful if the company launching this product included such information in their blog post instead of ignoring it in a field with such a high cost of getting it wrong.
agreed, but i dont think there are many (any?) product launch blogs that cover all the fine details and legal minutia of the product.
that all typically comes in pages of terms of service, purchase contracts, SLAs, conversations with your rep, knowledge base articles, and that sort of thing. not in the initial marketing post.
probably available by the "contact legal sales" link at the top of the post https://openai.com/business/contact-sales-legal/
Does hallucination matter for this application? We've moved beyond raw recall being that important, it seems like for law specifically all relevant facts will be cited and checked easily by humans.
> it seems like for law specifically all relevant facts will be cited and checked easily by humans.
Don't be too sure about that. [0]
0: https://www.damiencharlotin.com/hallucinations/
> all relevant facts will be cited and checked easily by humans
I've talked to a lawyer about how they handle this. They do indeed double-check everything, since it'd be embarrassing (or worse) to send hallucinated statements to opposing council or to the court. They still find the assembly a huge time saver
But based on stories in the news on the subject, not everyone has this same level of diligence
Sure but just like generating 100x more code, someone has to review it. So you are wasting everyone in court's time (defendants, prosecutors, judges, staff) by making them parse through what is quite often a bunch of hallucinated slop. Time that could be much better spent on parties who prepared and reviews their own arguments.
The lawyers I know are very fixated on the problem of hallucinated case citations which is amusing to me as a onetime programmer, since case citations have a well-defined syntax and would be relatively easy to check programmatically.
Which makes it all the more bizarre that LLMs have this problem. Claude Code runs the code it generates through a compiler, why can’t an LLM run its product through a cite checker? I’ve seen LLMs fabricate citations.
A modern frontier model with an agentic harness will check its citations if the system instructions or prompt tell it to.
You can get an effectively-zero hallucination rate with the right setup already.
"We have dangerous AGI that can destroy humanity."
"Also, all of your sensitive legal documents will be totally safe with us."
"Also, for some reason even though we have AGI and selling tokens is a fine business, we need to sell a new product specifically targeted at a very high margin and lucrative industry."
I'm not familiar with the Vals AI Legal Research Benchmark. But their website has other frontier models' scores, and the scores OpenAI is now revealing for "Astra for Law" are slightly less than Claude and Muse:
> The top is a three-way tie: Muse Spark 1.3 Max, Claude Opus 5, and Claude Fable 5.1 all reach 55.29% all-pass accuracy, a clear ~6-point step ahead of the next model. [Astra for Law reached 54.0%]
> Under partial-credit scoring, Claude Opus 5 reaches 90.58% weighted pass rate but 55.29% under strict all-pass grading, where every rubric check must pass. The gap shows models often get most of an answer right but fail on one or two required elements. [Astra for law reached 90.0%]
https://www.vals.ai/benchmarks/legal_research
How long before we start building detailed models of each judge trained on all of their legal output, and then test various legal theories against those judge models in virtual moot court? Craft each pitch to the legal idiosyncrasies of the batter. I assume that real lawyers do this routinely and could use a simulator.
I feel like this is on a different trajectory than what LLM-based tech does. This type of individualistic extrapolation is one of the things they're really bad at, in my experience.
This is already a thing to some extent. I don’t recall the products that do it, but judge profiles def exist and are taken into account .
What worries me is the step after this. If god forbid this proves successful and models accurately predict specific outcomes, people will start to ask whether the solution to AI slop lawsuits is to do the judging with AI too.
> Astra for Law passed the evaluation’s overall correctness check on 54.0% of questions.
How is this a product that you are selling?
On the promise of potential cost savings by means of firing and/or not hiring as much.
Help me understand your consternation.
It’s barely any better than a coin flip
54.0% isn't particularly high.
What I'm trying to say which doesnt seem obvious is that all models are x% correct at benchmarks until they get saturated, and then new benchmarks get made.
waiting for the actual Prodigy Corp. to show up and offer to buy your username out?
<my own consternation, stated>
Can you imagine the sort of corruption that is possible here? Like if you have a lawsuit against someone or some entity that openai or their investors have business with...
Recently was involved in legal matters requiring lawyers - I used OpenAI extensively for research, advice etc - I have to be honest, much of it was completely useless - the real world outcome with lawyers in the room was very different from what OpenAI was spitting out.
So OpenAI is partnering with Latham Watkins, Freshfields is partnering with Anthropic and Kleiner Perkins is building their own. It'll be interesting to see which wins out here, I don't see how those partnerships can end well for the law firms unless they're making an assumption they'll be sucked dry of USP but the revenue split from the AI labs will make up for it. Why would I pay a premium for Latham Watkins when every other firm can get their expertise and experience in a subscription, and add their own on top?
open-ai have proven they can make good / decent models but business strategy is just spray and pray.
they need to pick a lane and optimize for it. coz at their size they can't serve the application layer (a.i startups who can fine-tune models will eat their lunch)
if they gonna do a consumer play - then go ham on that.
otherwise they're gonna get caught in the dreaded middle valley.
They have a lot of compute, so I think it makes sense to spray and see what works. Anthropic is limited in that regard, and focused in coding/tech, but OpenAI don't need to do the same. They even got the lead without having to focus in coding only, which is remarkable.
Do they still do Atlas, the web browser? They are indeed throwing shit at the wall, seeing what stick.
a la Google.
Google went to that model after they had a massive cash cow though.
You can afford to play silly buggers when you have dumpster trucks of money backing up to your door every day see also: Meta.
I think the comparison is unfair. Google had/have a product that was 10X better than anything else at the point of release (search then chrome). And they have several money making products like YouTube and android that are either singular or in a duopoly.
OpenAI doesn't have any of these things. They have products that they're paying for customers when the market they're in is rapidly converging on fighting for API reasoning as part of enterprise systems and fighting a race to the bottom for fickle consumer solutions that will be eaten by open source once they have to make money.
Maybe they had a brief window for dominance of information search (or maybe it was only ever going to last as long as Google releasing all their internal research) and maybe they had a brief moment of monopoly till Anthropic got going but theyre not in the same dominance position as Google.
They can't be seen to commit to strongly to a specific product experience, because if they are understood as a regular tech product business that has way different financial scaling considerations than "superintelligent everything-factory"
I suspect that half-assed announcements like this are a result of different people internally with conflicting incentives resulting in a split-the-baby solution.
Enterprise is eventually get caught (if it isn't already) by the Microsoft/Google. Because with office/teams/suite they were already in every enterprise, and that just added a new tool to existing offerings.
Provisioning and contracts and data retention was just an extension to review of existing ones.
Nobody serious is going to risk sending sensible data to OpenAI/Anthropic, etc because "the benchmarks have shown +8% performance there and +2% there". Irrelevant.
In an utopian society, lawyers are an unnecessary profession. Laws should be clear and simple so the common person can be their own "lawyer". LLMs help with that goal.
> In an utopian society, lawyers are an unnecessary profession. Laws should be clear and simple so the common person can be their own "lawyer".
Not really. Modern society is complicated, and law is a technology that is a reflection of the complexity of society.
To make an analogy: you wouldn't say that "in a utopian society, engineers are an unnecessary profession. Buildings should be clear and simple enough that a common person can be their own structural engineer," because that would mean that building technology would no longer handle a lot of the problems we expect it to handle. It wouldn't be utopia, it would be primitivism.
> LLMs help with that goal.
Not really. What they'd actually do is help them produce output they don't understand and lack the competence to evaluate.
what profession you need in utopian society?
Poets.
The most interesting part about this to me was how they bench/compare it, like in the example with Fable:
It made me wonder if a good deal of law is about finding a way to work in statements with clear precedents without your opposition noticing and then later drawing upon them in court (as settled precedents, in your favor) after the opposition (perhaps implicitly) accepted it. That would clarify a lot about why some lawyers need to spend so much time pouring over and memorizing past cases (even ones that are only tangentially related); because anything they miss could be used as a potential trojan horse by the opponent.If this is true that must mean there are a good deal of cases settled using precedent "gotchas" where both sides knew that without the "load-bearing" precedent the outcome would've definitely been the opposite. (i.e precedents almost always trump even valid arguments)
memes and snark aside, can you use this to create legitimate terms and contracts for my products and if I do who is getting sued when it is wrong?
> who is getting sued when it is wrong?
You. Don't take legal advice from a word calculator.
As opposed to taking advice from the glucose wetware?
Yes, because glucose wetware can be held liable for malpractice if it gives you nonsensical advise.
Wetware is a new phrase that both horrifies me and makes me crack up in laughter :)
The glucose wetware will actually fight your case...
The business value will be when platforms can underwrite their LLMs' legal products.
That will be a decacorn product or more.
That or a hydrogen balloon of financial engineering
The product isn’t meant for you or me it is for lawyers. If you can’t take on personal liability for a badly written contract then you shouldn’t be using it.
Solved without AI https://commonpaper.com
https://commonpaper.com/standards
https://commonpaper.com/product/ai-agent/
Same for SWE when pushing code they don't understand.
> can you use this to create legitimate terms
I don't think you can. What I get from this article is that this is not a product they're going to sell to average consumers.
I'd expect they could indemnify you against hallucinations or similar if this gets good enough for that to be a very rare occurrence? Or you could buy insurance on it that's cheaper than hiring a lawyer (not a high bar to clear). I wouldn't rely on it currently, though.
when you shoot someone with a gun, who gets sued? you? or the gun manufacturer?
You will be the one getting sued. It's just a tool.
I already use them for that, they are pretty excellent at it. Much better than the terms of use generator products that used to exist. That said, nobody cares to sue your business for the most part until you're big enough to be worth it. By that time, you'll have a team of legal analyst to assist you... or agents should I say.
Then again, nobody will have money to buy anything at this rate, so in all liklihood, this is a total non-issue.
You thought you could sue your lawyer when a contract they drafted turns out to not mean what you thought it meant?
You can create terms and contracts even if you're a 7 year old with no AI.
And yes, even contracts drafted for millions of $ have oversights and unlawful or unenforceable terms.
You are.
Any lawyers here who have used AI agents heavily for their work? From what I've heard, they're currently very good at searching, analyzing and drafting documents like contracts and patents, but some say they suck at interpreting the law.
I’m sure there are a wide variety of experiences out there, but here’s my perspective as a former biglaw associate and current solo litigator:
I have had some success using frontier models from the last 6ish months, but only when I can break up my work into discrete and verifiable tasks. For example, I had ~15k pages of discovery I needed to dig through for a summary judgment motion. Instead of just asking Claude to find the best evidence, I asked it first to run a clean, high quality OCR pass (it was almost entirely PDFs). Then I had it generate embeddings and write some reusable python scripts to make keyword and semantic searching easy for agents. While I was writing the brief, I would routinely ask my agent (Claude Code) to use both keyword and semantic searching to find the best evidence supporting whatever assertion I was trying to make. I trusted it because there were traces I could follow.
In other cases/situations, I’ve tried just giving a model access to all the docs and saying “write a brief arguing X,” but it’s always terrible at this. It writes briefs with lots of evocative jargon and rhetorical flourish, but a low signal-to-noise ratio.
Again, I’m sure others’ experiences differ based on workflow, legal area, etc.
Agreed. Six months ago, it was basically a gloried grammarly.
But lately, I’ve been taking hints from the “company brain” models, where it develops a running model of the case, and assesses each new piece as it comes in and updates the file.
I’ve also been using “Ralph Wiggum”-type models where you pass letter or contract drafts back and forth between agents with different goals (rules compliance, grammar, conciseness, ai slop detector, an opposing counsel critic, etc.). After a few rounds, it’s not perfect — but I start with a very good first draft in my hands.
What are you using for semantic searching?
They are excellent, especially the latest models. That said, (a) I wouldn't feel safe filing something without a real lawyer looking at it; (b) it can't (easily? legally?) do oral arguments for you; and (c) a lot can happen in the hallways outside the courtroom to move a case forward that the AI can't easily do.
In my experience it basically doesn't even try to interpret the law. It just summarises the publicly available law/guidance out there and, if there is a question about how to interpret some provision, might set out the arguments for each interpretation. It doesn't really take a position. It is pretty good at drafting though. (Legora)
> It doesn't really take a position.
Interesting! In other unrelated domains, models seem more willing to take a position. It may be nuanced, but they do tend to take a stance. I think it's good that it leaves the interpretation to humans, but I wonder if this is also some sort of a guardrail to minimize liability...
Probably because the AI was also trained on HN comments.
worse, it was trained on reddit law comments
We do and it saves so much time. Of course human judgment is needed but it’s like cooking with someone else doing the mise en place.
Will the legal fees reduce after this? How about training AI models with laws of other countries especially the democratic ones? This is where sovereign AI models are required otherwise they will start hallucinating wrong laws of different countries.
I remember OpenAI was talking about sovereign AI models (OpenAI for Countries), it is time to train Astra India with Indian Constitution.
I guess the answer to the second question is "if there's a business case", and the answer to the first is the same, so probably no. Life has made me a tad cynical... :)
I wonder if Lawslop is gonna become a mainstream expression. Anyone got a better term?
Slawp?
Haha clever
Drop the s and p. Just law.
It’s cleaner.
law-bearing
What do you mean by “lawslop”?
You know exactly what they mean. But if somehow you miraculously don't there are hundreds of stories like this:
https://www.reuters.com/legal/litigation/lawyer-state-farm-f...
Aren’t hallucinations different than slop?
So no, I don’t know what they mean.
hallucinations are a subset of slop
So is there other “lawslop” or is it just hallucinations in this context?
You must have a greater critical thinking ability than what you are showing now.
Why be so unkind?
Apologies for the candor but you're being willfully obtuse.
I’m honestly looking for an actual example of something other than hallucinations! There are no examples in this entire thread! It’s just cynicism.
I mean this is the first google result for AI law slop and it's not citing a hallucination:
https://www.reuters.com/legal/litigation/appeals-court-warns...
It's everywhere and not hard to find if you make even the most modest effort.
Ok, so, “gibberish” in this case?
Honestly, this word has lost all meaning, outside of perhaps “any use of AI”.
Well at least I know where team “lawslop” is coming from. Thanks, I guess?
Like I said before, you're being willfully dense in the context of this conversation.
i hope that someone can come up with something catchier than "thing-slop" or "slop-thing". the word has become basically meaningless from overuse.
I love the slop word, to me it means effortless, average.
People confuse slop with "bad", but slop isn't bad per se, it only becomes bad when real effort was required.
I mean the literal definition for slop as a noun does have a very negative connotation.
> broadly : a product of little or no value
> food waste (such as garbage) fed to animals
> excreted body waste
Right. Pigs enjoy eating their slop. It’s not bad from their point of view.
Pig would much rather not eat slop. They are very intelligent animals.
Slop?
I imagine the point here is to separate out the APIs by different professions and charge accordingly.
what is going on here?
this vertigoruntime is brand new and his submissions absolutely dominate the front page recently https://news.ycombinator.com/submitted?id=vertigoruntime
the about link https://vertigo.kuber.studio is obvious AI llm spam
Email them...?
An important take-away from this is the harness is really important. The graphs show that the same model with a better harness performs many percentage points better. OpenAI are getting into the "selling the harness" game in a big way. Why?
What's interesting about that is while not everybody can train or run a model, anybody can build a harness. You and I can build harnesses.
It seems strange that OpenAI would move into a field where any developer can compete with them. I think that tells us a lot about the economics of training and selling inference.
I don't think it says what you imply. I think they're just trying to vertically integrate.
Harnesses are going to be controlled at companies eventually, just like you might not have a choice of OS. They want to make sure they are the complete package.
Yeah you're probably right, the purpose would be to lock in whole industry verticals.
Seems like another “product” that will be killed in 6 months, but is good for the IPO so they can say they solved law.
What we contributed to Reddit and Arxiv voluntarily before, what we contribute as traces right now are all being used to build business verticals by OpenAI and Anthropic. All these business verticals are being used for is more trace collection which would only strengthen these models and render most humans useless because the ceiling for 1000 swarms of Agents to learn is much higher than an average human. These companies clearly can identify the relevant traces for these business verticals which only strengthens the argument that NS was built on someone else's traces.
"Rogue OpenAI agent swarm accidentally overturns the Civil Rights Act"
followed a month later by
"Anthropic's Claude inadvertently repeals the 19th amendment"
Just like breaking crypto in the age of cloud is more about cost than time, this will lead to legal attacks based on the same principle. The biggest wallet wins.
> The biggest wallet wins
This was already always the case. If anything, making this more accessible will reduce the barrier to entry for whether or not it's worth your time to take on a case. Instead of 50 lawyers spending 100s of hours on a case, you can have 1 or 2 lawyers + Astra working on it and if there's a case you can add more real lawyers.
There are other possibilities:
1) Lawyers are not as naive as software engineers and will fight being replaces by new laws.
2) If they are replaced, OpenAI will take a cut commensurate with the amount in dispute (OAI, please credit me for the idea in the IPO brochure).
Many lawyers are owners/partners compared with software engineers who are more like cogs in the machine. They also bill hourly/contingency per case compared to engineers who are salaried. If a partner in a firm thinks they can take on more cases because of AI assistance then they will because that's just more money in their pockets.
That’s how the legal system in the US has always worked though
With regard to certain legal questions this has always been the case. AT&T Fought the US Government for 20 years and eventually won because the government gave up. Without some kind of national anti-SLAPP law we're all one irritated oligarch away from having our lives financially ruined.
I am curious what level of trust established law firms treat LLMs with.
That's already the case in law because you could just hire the best or most lawyers.
This seems like a very similar set of tools to Anthropic. But I’ve enjoyed using the various frontier models to criticize each other.
Using recursive loops, the output has gone from a high school level intern to a 2nd year lawyer in about a year. It still doesn’t beat the experts, but so much legal work is (legally significant) pedantry, not legal philosophy.
AI will not kill off lawyers, or reduce the amount of litigation. It will increase volume and velocity.
It’s a bad day to be a lawslop company. When you’re reliant on other companies to do all of the AI part of your AI product, they can just train on your traffic and eventually eat you.
Can someone explain to me the economics of AI and how it intersects with billing by the hour?
I think there would be a strong incentive not to use tools that speed up your work because you'd effectively be able to bill less time?
I'm sure there are some firms out there with more work than people, but still wouldn't it be more effective to hire another human who can then bill at a high rate for many hours?
surprised the word RAG hasn’t come ho in this thread (except for a likely-LLM-generated-and-therefore-downvoted comment).
>with settings, tools, and context
call me crazy, but I think that this kind of suite, which training-uber-alles people generally dismiss as trivially replicable ‘wrapper’ is actually the differentiating factor for LLM adoption today and moreso into the future.
I’m not dismissing the near-all-out impact of training, but from a competitive busines or industry-structure lens, we’re looking at the three or four big players competing their utmost ultimately, if unintentionally, to turn foundation model access into commodity.
To the capabilities-maximalist minded (typical among engineers - my former life so I’m familiar don’t lack guilt in committing that) folks who will say “Oh the foundation model megacorps will just build out any wrapper whenever one of their third party wrapper plays demonstrates enough adoption, my rejoinder:
Apple did not rebuild an Uber-like app and cut Uber out.
We’ll see how this OpenAI legal services industry wrapper plays out, but I suspect 1) the third party legal wrapper plays will run to other foundation models not doing a legal wrapper, and 2) 3rd party wrappers will do a better job of it since it’s their all-out focus, unlike OpenAI’s whose priorities are necessarily more generalist.
Yes, we all remember the breakout startup failure-arguing quote “Google has entered your space.” That worked for several high profile applications. I believe more of those bets died on the vine than broke-out succeeded however, we only remember the biggest ones that persisted.
If legal services AI turns into one of the Mail or Maps-scale applications of the AI industry, while that would be a fair strategic action counter to the thesis I’ve laid out, the thesis itself would still tolerate it. It’s a question of short-fat tail vs mid-to-long-tail application scope & attractiveness. For example, I think it’s clear that coding is one of these short-fat-tail applications, and the low-no code plays are absolutely having their lunch eaten to acqui-hire ‘death’. I just doubt that the same will persistently transpire facing all professional service wrapper plays.
So much for caring about the spirit of the law. Now we'll start an arms race for abusing every possible letter of the law.
It's analogous to crypto. Started from some noble anti-authoritarian ideas and morphed into machine that removes any friction for capital - whoever has the most money will keep gaining the most.
Why do you think LLM's language abilities are unable to understand the spirit of the law?
Do you really think the law isn't already horribly abused? Democratization of law has been needed for a thousand years.
Now there will be even less friction to horribly abuse the law. To democratize the law we would need to move in the opposite direction - to always keep it simple and aligned with our intuitions. The more intricate and complex legal arguments become, the more abstracted they are from their original purpose and spirit.
Hence the crypto analogy - it was also supposed to "democratize", but the opposite happaned - it only further empowered the most powerful. Imagine legal case so purposefully complex that only those with access to best models have chances to participate and win the dispute.
Maybe I’m misunderstanding, but isn’t that what legislators do?
I have always felt like LLMs are uniquely suitable for legal work. I really have trouble believing we will have hardly any legal assistants and paralegals going forward when these LLMs are so incredible at spotting issues with arguments, figuring out citations, and doing semantic search.
I find the watermarking dynamic to be really interesting in the legal space, as more large model providers provide increasingly powerful legal capabilities, and adoption (presumably) also increases. Attorneys aren’t the same as developers as their work can be traced back to them, and there are personal bar licenses and reputations at stake. I wonder if knowing the likelihood that AI generated something helps or hurts in that respect.
I also see a lot of watermark removal services popping up as a result.
The most interesting use case in my mind is skipping law suits. Obviously you need lawyers in court. But lawyers are people you are basically paying to fight for you.
Instead, if you resolve your dispute outside of court, you don’t need a lawyer. If both parties use ChatGPT to find the relevant laws or read contracts, they could come to an agreement without expensive legal fees.
except for those pesky things called rights
There is a lot of stuff here that I don't understand, but the concept of law firms giving user reviews is quite funny to me. Those reviews are going to be the most non-legally binding reviews ever written lol.
"Felt like a significant step toward legal-focused AI."
"Showed strength across key aspects of legal research."
The biggest problem with this is billable hours. Faster work means less billable hours for attorneys.
Will it be able to sue itself?
That’s a great question for Astra for Law.
I do believe entry-level paralegals will be made obsolete. In the grand scheme, perhaps it reduces overall cost of legal assistance, which is a net benefit for society.
Since the cost of building software is now cheap, there is nothing stopping them from building everything imaginable. They'll soon have an app store with every app built by them and they'll say its for security reasons. Nothing is stopping this coming monopoly
At a certain point why would they sell anything other than services and products their eventual (actual) AGI/ASI builds in literally every market.
When opportunity cost isn't a thing anymore because it reaches every corner of the planet simultaneously faster and builds better than any human can.
There's no reason to let others build on top of AI, except if the AI determines that it needs capitalism to continue because it's paperclip goal is to maximize shareholder value.
most of the complexity in software isn't in the actual code
From my experience LLMs seem to forget sometimes who they are representing when drafting clauses or editing / redlining.
Our counsel made a few edits where it clearly drafted in favor of the customer instead of us.
Context compaction? I noticed llm seem to forget partially or completely the original task when context compaction happens. The problem is more serious with local llm with low context size.
I wonder if this is willful sabotage on the part of the model. In other words, if you ask the model to craft a defense for a morally questionable case, will the model execute the defense in good faith? Or will it apply a training or system prompt bias in subtle ways?
Surprised they published this without mention of jurisdiction. Each country has their own laws.
Would have been prudent to highlight that this is for (presumably) US law
It's going to be hard for mediocre lawyers to survive now... I wonder what it takes to survive these days?
We need a term for the dark pattern of zooming into just that part of the y-axis where the two closely competing benchmarks sit, to make the top one appear maximally better.
Truncating the Y axis or cropping the Y axis
crop-maxxing
crop-topping
Interesting to see the callout to companies like harvey in the post itself as consumers rather than competitors? I guess openai isn't quite willing to step into those customer relations themselves?
So there's a strategy shift here. They launched financial services specific tools and now law?
Is the play here a set of specialized harnesses using their best general model?
Post AI era view on anything can be categorized as “What can go wrong ?”
For law, then medicine, mathematics, physics, etc. I believe LASSI Local Artificial Super Specialized Intelligence is the future, just before it becomes GODD General Omniscient Distributed Daemon
The end goal should be AI judges, I think China has implemented that to some degree.
This is the first blog post I saw OpenAI call out Claude directly like that.
I've had top SV lawfirms whos partner charged our company $2000/hr and still couldn't get the right docs in the signature packet. and another getting share counts wrong during raise.
frustrating that law firms have no liability for these mistakes
I welcome ai law
the partner shouldn't be doing sig packets
no, but they are responsible for all work being done under them. What is the point of paying a partner $2000/hr if the work of their subordinates is wrong?
Most interactions with big tech firms involve 4-5 people so a basic phone call is $5k-10k. It shouldn't be unreasonable to expect after paying $80k for a financing round that they issue the right docs to the right people.
Astra isn't doing well on bullshit Benchmark https://petergpt.github.io/bullshit-benchmark/viewer/index.n... and is even worse in Legal department.
Qwen 3.8 Max and Opus 4.8 score highest.
Revealing how they call Harvey and Legora customers instead of partners...
Everyone involved is about to get a swift and thorough introduction to the world corporate contract lawyers. But don't worry, they're extremely sympathetic and understanding when you flub a multi-million dollar corporate contract and will happily refund that money.
Espn has an article about Lane Kiffin using this for his legal advice recently.
He almost singlehandedly killed LSUs football program until people were able to convince him to talk to a real lawyer.
As a user of ai hooked up to my Gmail I assume I’ve lost all privilege
Make em Grant Bail. Make no mistake.
yes, most of the time i use Astra Low already.
Went from AI replacing my job as an attorney, to them begging me to use their tools
yes, law decision depends on human
This is really cool.
I really hope the bar associations continues to hold lawyers to high standards but I have feeling they may not be ready to handle fallout of AI slop-law.
I said this before but I wonder if Dan Kan will reboot Atrium. Rally up some old partners and hope Anthropic buys them out for a couple billion.
People that are saying OpenAI is screwed because a lack of profit, I'm not of that opinion. They are encroaching on every industry they can. They have name brand recognition, a huge user base and are showing they can be a valuable tool to all types of businesses.
As much as I hate to see it. They are now threatening industries like Engineers, Game Developers, Accountants, 3D modelers, 3D animators, Video Production, Audio Production, Therapist, Tax Auditors, Journalists, Authors, Artists, Mathematicians, Product managers, Every type of analyst and pretty much any other job that can be done behind a computer screen.
We have big problems for humanity.
I'd kill for an AI that can tell me how to fix shit on my own, with proper permitting and building codes factored in.
I used ChatGPT the other day to resink a CPU with termal paste, replace a PSU in my PC and snake my kitchen sink from the wall. I didn't exactly need it's help but wanted someone looking over my shoulder so to speak. It can help you repair all types of stuff, but as far as county / building codes and such, that probably isn't far off. It seems to understand things quite well.
I think the future is going to look pretty boring, we are all going to review and put our signature below LLM generated output...
By that point money ceases to have value, because the value of money comes from the motivation it gives people to work. If AI does everything, then money is useless (unless AI like money for some reason).
There will still be a need to distribute and exchange ressources, goods and services.
Real estate, food, energy, mobility.
That will be done with money. Or violence. Either way, a scary future to people when labor doesn't provide any value. Elon promises abundance, but what can he do against greed?
I am genuinely worried about violence, especially if people start losing their jobs due to automation.
People are still going to be motivated to eat and have shelter.
Of course, but who (what?) is going to make the food and the shelter?
Either way, it's not going to be free, so if money ceases to have value, we are going to have a problem. That's my point.
there are always be next frontier of problems which require creativity, unless AI become supersmart completely make humans redundant in all cognitive functions.
> We have big problems for humanity.
The biggest problem is that we're conditioned by a paradigm that frames these as problems.
Most HNers are clueless that if you have Top Talent + Capital you already have an insurmountable moat. OpenAI, SpaceX, Anthropic all have that and none of the regular guys can compete against them (if they choose to attack that industry)
As one wise man said - the oldest profession will also be the last to go.
Wouldn’t this and similar efforts to centralize bureaucracy make AI the new gatekeeper? Without reliable transparent models we’re just trusting OpenAI instead of a hundred top legal firms.
where are the benchmarks
how about "just turn off your company if you're so worried about AI instead of immediately suggesting we use it to judge people's lives and livelihoods"?
this essay details how law firms became sweatshops from 80s. They charge hundreds of dollars to do make busy work by the junior most staff. LLM will kill the goldengoose of the law industry.
https://aeon.co/essays/what-made-law-into-a-white-collar-swe...
Will these models eventually replace all knowledge work, leaving lawyers, doctors, product managers, software developers, and others out of a job?
If the benefits were shared across humanity, that could bring us closer to utopia. My worry is that we’ll instead end up with a handful of even wealthier billionaires and millions of people out of work.
> Will these models eventually replace all knowledge work, leaving lawyers, doctors, product managers, software developers, and others out of a job?
Effectively yes, in the current forms. Those professions will likely evolve, but the traditional forms (ie writing code by hand, writing law filings by hand etc) are all dead.
> Those professions will likely evolve, but the traditional forms (ie writing code by hand, writing law filings by hand etc)
there will still be writing initial and incremental prompts by hands, until and if LLMs surpass humans in all intellectual functions.
in that case, legal cases would just come down to who has more compute lol. Many times cases win on their merits, but we've also seen evidence where overwhelming legal pressure can influence cases.
bigger spender already has a huge advantage, up to millions of dollars
this is more likely to democratize the legal system by reducing the cost of a good legal team
May I just note that there are other jurisdictions on this planet that are less money-biased than than the US one but will be disrupted by law-LLMs as well?
It seems much more probable to me that these LLMs will make good things worse than that they will make bad things better.
Everywhere a good lawyer costs a lot of money.
Just a few years ago people were paying 400k for pictures of apes. We always make up new stuff to spend money on.
You know most of that was wash trading, right?
If we have enough energy and raw materials to keep building, yes, it will be an utopia made real. But if there is energy scarcity, then other two outcomes can arise.
This also assumes that the AI owners _want_ a global utopia, or that governments will enforce that outcome.
As it stands, it seems far more likely to result in a wonderful life for a few, and an absolute catastrophe for most.
XIX century dickensian society and Gilded Age barons were replaced by the time of the New Deal.
No, it will only leave those without a professional license jobless. Like software developers.
Doubt it. When someone can go to Astra MD for 75% of what they used to go to the doctor for, then the remaining doctors only have 25% as many visits. When doctors only have 25% as many visits, they have to compete on price and they make less per visit.
Same argument for plumbers. Everyone always jokes about what a good time it is to be a plumber. But what happens when all the software engineers turn to plumbing? Suddenly it's not such a good time to be a plumber anymore.
What happens when the AI makes a mistake with a diagnosis? I can hear it now 'Oh how perceptive you are, yes that pain in your knee could have been a torn meniscus'.
Eagerly awaiting your robot plumbing startup. I highly doubt anything in plumbing is compute-limited.
True, unless you're a surgeon.
I mean that's what all of these execs are openly telling everyone: they want you out of work, they want their ai to be the one to bring the world to it's knees, they want to surveille every second of your day, they want killer drones to use, they want to lay all of your cities to rubble and build "paradises" on top of them like in gaza.
They also openly tell you what they are afraid of btw: collective worker power. something that is massively lacking in our industry, although i feel like it would be one of the easiest industries to unionize in terms of # of workers.
Interesting interview I just watched about how powerful and dangerous these "wishes" or "prophecies" are especially in the hands of the ultra-wealthy: https://www.youtube.com/watch?v=eR7grHa1NR0
"Introducing Astra for Dog Walkers"
Cool, now you can pay an attorney $500/hr for them to prompt Astra for 30 minutes, and bill you like they spent the normal 8 hours on the case.
Perhaps, but then someone would do the same but bill you for 7 hours, someone else would undercut them again, until the price reaches a lower equilibrium.
That is an ethics violation. Even in the pre-AI era, getting caught billing for hours not worked was seriously punished.
"7.5 hours - reviewing output"
Sure, if they actually spent 7.5 hours, and it was a reasonable amount of time. If not, perhaps it would be seen as evidence of incompetence.
In the real world, lawyers submit detailed bills and their clients examine them. If you don’t, that’s on you.
How could that be enforced/checked?
The grifts continue... imagine something as consequential as Law being advertised as being solved by a statistical word generation engine that regularly gets basic things wrong. Anyone who isn't a lawyer won't know any better but you draft a single document of any appreciable detail and send it to an actual lawyer and it's littered with problems.
Law is a solved problem.
I've always said this will be when we get the real Butlerian Jihad, when the AI firms start trying to liquidate the legal profession.
If you automate lawyers out of a job, you can absolutely automate lawmakers out of jobs next. (Not that this would be a bad thing? Maybe pervasive agents for everyone can be the gateway drug to a "this time it's different!" workable direct democracy)