I would love to see these benchmarks split up across "generic" tasks and novel tasks.
For example, redlining an OEM agreement with my vendor. They pretty much all use the same template, and the objections are always the same. I tried using Claude first, and even my lawyer was impressed that it found things they missed. Then I used Claude to redline that type of contract and the other side said "your lawyer was surprisingly thorough". I had my lawyer do the final check, but saved a bunch of money doing the first three rounds with Claude instead. And my lawyer basically just said "looks good to me" by the time they reviewed it.
Or even writing a simple will or estate document.
It seems like LLMs are great for the kind of generic work you'd usually pawn off to a new associate or paralegal, but not for the important things that partners usually do.
show comments
mlmonkey
I might be incredibly naive here, but I really think LLMs can make a difference in _some_ areas of law.
Take, for example, India. The courts are backlogged like crazy. If you're not willing to bribe a judge or the opposite side's lawyer, you'r looking at multiple decades to settle simple cases.
I wish I could waive a wand and have some LLM just go through all of the 58 million pending court cases, identify the clear cut ones and free up the judges' time so they can work on the more difficult ones.
How could we trust legal product from a company who was ignoring privacy protection law like that?
show comments
victor9000
I’m not sure if this is the right solution, but we clearly need better tools for the legal space. Anthropic's recent policy update makes this abundantly clear:
“When our products are used in ways that can affect someone's health, legal rights, finances, livelihood, or access to essential services, we require there to be a qualified ‘human in the loop’ (that is, someone who has the authority to review and, if necessary, change Claude’s recommendations), and for the individual affected to be told that AI was used.”
Something as common as a family law case has the potential to affect all of the above. Now imagine being a petitioner, respondent, paralegal, or other 3rd party using Claude to review a draft document only to find out that Anthropic has sent your entire file to opposing counsel.
It sounds egregious, but Anthropic is clearly stating they will use their own judgment to control what people can do with its product, and that they will treat customer data as if it's theirs to disseminate. So there's a massive opportunity for an unbiased tool willing to fill this void.
show comments
otterley
As an attorney, I am not particularly impressed about a 20% pass rate on a benchmark. The name of the game for litigation practice is about being prepared for novel situations, not relitigating past ones.
show comments
mathis-l
Not sure how good the Harvey benchmark is. I recently looked into traces from an agent solving their firm knowledge task subset and found multiple problematic tasks and rubrics.
As somebody who is forced to deal with lazy attorneys, I am thrilled that LLMs are starting to push in this space.
First it is going to suck, like it did with coding but after a while it is going to be good enough for 90% of legal work. And finally when attorneys will feel the competition of AI doing their work for 10 cents of tokebs, they are going to be less arrogant and more motivated to finish a task instead of client babysitting them through the process.
Havoc
Law is country specific so I'm still somewhat confused about how anyone would trust this.
If I ask it about Irish law is it gonna mix in some US and British law in there because the training data has lots of that? Or hallucinate a mixture?
show comments
postalcoder
Looks like this is not an open model. I'm curious how the sentiment toward chinese models will shift as labs trend toward keeping them proprietary.
Also, I wonder if this is at all related to the recent reports that Harvey has been getting squeezed on margins by the US labs.
"Freakazoid has ... access to all of the Internet's knowledge. He cannot fly, but often runs around with his arms extended and making whooshing sounds with his mouth, pretending he is."
show comments
Trusteando
The non-disclosure of the training set may raise legitimate skepticism about the training set containing tasks similar to those in the test set. That would make the benchmark result moot. There are benchmarks that by design can't be manipulated by incorporating similar tasks in the training set, for example those that dynamically generate tests and only reveal the global result. Testing on one of those would give a strong measure of the strength of this legal AI framework.
toephu2
Sort of ironic, in Google Chrome, when attempting to visit that legal z.ai site I get:
Your connection is not private
Attackers might be trying to steal your information from z.ai (for example, passwords, messages, or credit cards). Learn more about this warning
net::ERR_CERT_COMMON_NAME_INVALID
show comments
Depurator
Curious what domain is next, GLM-Science? And when will it be profitable to drill down to even more domain specific models like GLM-Economic-Geography or GPT-Aquaculture?
show comments
mattfrommars
I’ve been wanting to get into legal tech but I never managed to find the opportunity to get in or found folks who are interested in this space. Now I wonder if it’d bit too late to get into this space since now they are startups valued in billions
show comments
consumer451
Wait, so this not an open weight model, correct? Only available via z.ai?
So, if a US or EU person shares their private legal info here, it's processed and stored by an adversarial country that has no extradition treaty with their country?
show comments
Betelbuddy
No discussion on this can avoid mentioning this ongoing disaster...
A lot of the examples they show seem to be focused on common law countries. The US and UK are where most of the money is to be made in the legal industry, but I am curious to see what ripple effects we get from people attempting to use AI tools geared toward US/UK-style legal systems in other kinds of systems.
senshan
The usual concern is attorneys, but judges are probably even more overloaded these days. Can this help judges as well? If so, would this ultimately lead to better decisions?
show comments
zerozerotwo
I don’t even understand how a legal llm works in a place in China. Can it ever spit out a decisions against the party?
yipinwong
A progress for sure. Just don't expect GML-legal to answer about Xi or CCP for any US courts that mention those terms.
aliljet
How does this compare with existing frontier models especially when processing massive volumes of evidence that might be contradictory...
bb123
Would be nice if it said what legal system it is set up for. US law? Chinese law? European law? All of them? The author seems to assume we'd know.
show comments
alansaber
Video is very SAAS-demo coded. I wonder if they worked with any in-domain legal experts at all.
maxdo
No fable 5.1 no opus 5.5 , no gpt 6.1 , benchmarks for the hype .
Havoc
Unexpected direction for zai to be going into
theokrueger
the liability model of these models is poor, i think progress in that domain should be a heavy focus.
show comments
bigstrat2003
Given how bad LLMs are at the domain I'm an expert in (programming), it will be a cold day in hell before I trust them to give me legal advice (or medical advice). Those are not trivial things where it doesn't matter if you get it wrong.
lenerdenator
There should be laws about disclosure on the part of attorneys about their use of AI. I know they're already using it, but spending years or decades in prison because Lionel Hutz, Esq. did something stupid at a prompt sounds unappealing to me.
show comments
nullc
Great-- send your legal issues to a third party who will have to produce them in response to a subponea. :-/
Come on, z.al --- post the weights!
crossroadsguy
Lawyers, essentially the exorbitant and freeloading self-appointed gatekeepers of the deliberately arcane, unaffordable, inaccessible, and patently unequal legal systems all across the world, will now have even their routine/rote, repetitive, menial work done by cheaper LLMs for cheap.
show comments
alexnewman
Why don’t I lose all legal privs when I hook up my email to these third party services
newsclues
Professional AI is so interesting to me, legal, medical, engineering, accounting all can benefit.
And the opportunity for a moat of AI that is correct and accountable(legally) is ripe for the picking.
dyauspitr
Lawyers are going to start living gilded lives. I don’t see us getting past their professional human sign-off so they are essentially going to be guaranteed, easy, high paying jobs with the LLMs doing all the heavy lifting in the future.
show comments
Natsu
It looks like they're relying on tracked changes and human review, but I didn't see anything addressing something that has frequently made the news: LLMs outright hallucinating cases or holdings.
If the human lawyers don't catch that on review, they're likely to have an angry judge and an AI saying "absolutely, you were right to call me out on that, I shouldn't have hallucinated a non-existent case and made the judge sanction you."
iLoveOncall
"advancing". Absolutely laughable, we're talking about less than 20% accuracy on benchmarks. 100% useless.
I would love to see these benchmarks split up across "generic" tasks and novel tasks.
For example, redlining an OEM agreement with my vendor. They pretty much all use the same template, and the objections are always the same. I tried using Claude first, and even my lawyer was impressed that it found things they missed. Then I used Claude to redline that type of contract and the other side said "your lawyer was surprisingly thorough". I had my lawyer do the final check, but saved a bunch of money doing the first three rounds with Claude instead. And my lawyer basically just said "looks good to me" by the time they reviewed it.
Or even writing a simple will or estate document.
It seems like LLMs are great for the kind of generic work you'd usually pawn off to a new associate or paralegal, but not for the important things that partners usually do.
I might be incredibly naive here, but I really think LLMs can make a difference in _some_ areas of law.
Take, for example, India. The courts are backlogged like crazy. If you're not willing to bribe a judge or the opposite side's lawyer, you'r looking at multiple decades to settle simple cases.
I wish I could waive a wand and have some LLM just go through all of the 58 million pending court cases, identify the clear cut ones and free up the judges' time so they can work on the more difficult ones.
Seems like people forgot Z.ai was snapshotting the entire workspace before every prompt. https://blog.ferstar.org/en/posts/zcode-silent-workspace-sna...
How could we trust legal product from a company who was ignoring privacy protection law like that?
I’m not sure if this is the right solution, but we clearly need better tools for the legal space. Anthropic's recent policy update makes this abundantly clear:
“When our products are used in ways that can affect someone's health, legal rights, finances, livelihood, or access to essential services, we require there to be a qualified ‘human in the loop’ (that is, someone who has the authority to review and, if necessary, change Claude’s recommendations), and for the individual affected to be told that AI was used.”
Something as common as a family law case has the potential to affect all of the above. Now imagine being a petitioner, respondent, paralegal, or other 3rd party using Claude to review a draft document only to find out that Anthropic has sent your entire file to opposing counsel.
It sounds egregious, but Anthropic is clearly stating they will use their own judgment to control what people can do with its product, and that they will treat customer data as if it's theirs to disseminate. So there's a massive opportunity for an unbiased tool willing to fill this void.
As an attorney, I am not particularly impressed about a 20% pass rate on a benchmark. The name of the game for litigation practice is about being prepared for novel situations, not relitigating past ones.
Not sure how good the Harvey benchmark is. I recently looked into traces from an agent solving their firm knowledge task subset and found multiple problematic tasks and rubrics.
Here are 3 examples (there are more): https://github.com/harveyai/harvey-labs/issues/191
As somebody who is forced to deal with lazy attorneys, I am thrilled that LLMs are starting to push in this space.
First it is going to suck, like it did with coding but after a while it is going to be good enough for 90% of legal work. And finally when attorneys will feel the competition of AI doing their work for 10 cents of tokebs, they are going to be less arrogant and more motivated to finish a task instead of client babysitting them through the process.
Law is country specific so I'm still somewhat confused about how anyone would trust this.
If I ask it about Irish law is it gonna mix in some US and British law in there because the training data has lots of that? Or hallucinate a mixture?
Looks like this is not an open model. I'm curious how the sentiment toward chinese models will shift as labs trend toward keeping them proprietary.
Also, I wonder if this is at all related to the recent reports that Harvey has been getting squeezed on margins by the US labs.
https://quoteinvestigator.com/2019/07/30/lawyer/
"he who will be his own Counsellour, shall be sure to have a Fool for his Client."
How much worse the lawyer who subcontracts Freakazoid?
https://en.wikipedia.org/wiki/Freakazoid!
"Freakazoid has ... access to all of the Internet's knowledge. He cannot fly, but often runs around with his arms extended and making whooshing sounds with his mouth, pretending he is."
The non-disclosure of the training set may raise legitimate skepticism about the training set containing tasks similar to those in the test set. That would make the benchmark result moot. There are benchmarks that by design can't be manipulated by incorporating similar tasks in the training set, for example those that dynamically generate tests and only reveal the global result. Testing on one of those would give a strong measure of the strength of this legal AI framework.
Sort of ironic, in Google Chrome, when attempting to visit that legal z.ai site I get:
Your connection is not private Attackers might be trying to steal your information from z.ai (for example, passwords, messages, or credit cards). Learn more about this warning net::ERR_CERT_COMMON_NAME_INVALID
Curious what domain is next, GLM-Science? And when will it be profitable to drill down to even more domain specific models like GLM-Economic-Geography or GPT-Aquaculture?
I’ve been wanting to get into legal tech but I never managed to find the opportunity to get in or found folks who are interested in this space. Now I wonder if it’d bit too late to get into this space since now they are startups valued in billions
Wait, so this not an open weight model, correct? Only available via z.ai?
So, if a US or EU person shares their private legal info here, it's processed and stored by an adversarial country that has no extradition treaty with their country?
No discussion on this can avoid mentioning this ongoing disaster...
"AI Hallucination Cases" - https://www.damiencharlotin.com/hallucinations/
A lot of the examples they show seem to be focused on common law countries. The US and UK are where most of the money is to be made in the legal industry, but I am curious to see what ripple effects we get from people attempting to use AI tools geared toward US/UK-style legal systems in other kinds of systems.
The usual concern is attorneys, but judges are probably even more overloaded these days. Can this help judges as well? If so, would this ultimately lead to better decisions?
I don’t even understand how a legal llm works in a place in China. Can it ever spit out a decisions against the party?
A progress for sure. Just don't expect GML-legal to answer about Xi or CCP for any US courts that mention those terms.
How does this compare with existing frontier models especially when processing massive volumes of evidence that might be contradictory...
Would be nice if it said what legal system it is set up for. US law? Chinese law? European law? All of them? The author seems to assume we'd know.
Video is very SAAS-demo coded. I wonder if they worked with any in-domain legal experts at all.
No fable 5.1 no opus 5.5 , no gpt 6.1 , benchmarks for the hype .
Unexpected direction for zai to be going into
the liability model of these models is poor, i think progress in that domain should be a heavy focus.
Given how bad LLMs are at the domain I'm an expert in (programming), it will be a cold day in hell before I trust them to give me legal advice (or medical advice). Those are not trivial things where it doesn't matter if you get it wrong.
There should be laws about disclosure on the part of attorneys about their use of AI. I know they're already using it, but spending years or decades in prison because Lionel Hutz, Esq. did something stupid at a prompt sounds unappealing to me.
Great-- send your legal issues to a third party who will have to produce them in response to a subponea. :-/
Come on, z.al --- post the weights!
Lawyers, essentially the exorbitant and freeloading self-appointed gatekeepers of the deliberately arcane, unaffordable, inaccessible, and patently unequal legal systems all across the world, will now have even their routine/rote, repetitive, menial work done by cheaper LLMs for cheap.
Why don’t I lose all legal privs when I hook up my email to these third party services
Professional AI is so interesting to me, legal, medical, engineering, accounting all can benefit.
And the opportunity for a moat of AI that is correct and accountable(legally) is ripe for the picking.
Lawyers are going to start living gilded lives. I don’t see us getting past their professional human sign-off so they are essentially going to be guaranteed, easy, high paying jobs with the LLMs doing all the heavy lifting in the future.
It looks like they're relying on tracked changes and human review, but I didn't see anything addressing something that has frequently made the news: LLMs outright hallucinating cases or holdings.
If the human lawyers don't catch that on review, they're likely to have an angry judge and an AI saying "absolutely, you were right to call me out on that, I shouldn't have hallucinated a non-existent case and made the judge sanction you."
"advancing". Absolutely laughable, we're talking about less than 20% accuracy on benchmarks. 100% useless.