Lawyer Caught Using AI While Explaining to Court Why He Used AI (404media.co) 39
An anonymous reader shares a report: An attorney in a New York Supreme Court commercial case got caught using AI in his filings, and then got caught using AI again in the brief where he had to explain why he used AI, according to court documents filed earlier this month.
New York Supreme Court Judge Joel Cohen wrote in a decision granting the plaintiff's attorneys' request for sanctions that the defendant's counsel, Michael Fourte's law offices, not only submitted AI-hallucinated citations and quotations in the summary judgment brief that led to the filing of the plaintiff's motion for sanctions, but also included "multiple new AI-hallucinated citations and quotations" in the process of opposing the motion.
"In other words," the judge wrote, "counsel relied upon unvetted AI -- in his telling, via inadequately supervised colleagues -- to defend his use of unvetted AI."
The case itself centers on a dispute between family members and a defaulted loan. The details of the case involve a fairly run-of-the-mill domestic money beef, but Fourte's office allegedly using AI that generated fake citations, and then inserting nonexistent citations into the opposition brief, has become the bigger story.
New York Supreme Court Judge Joel Cohen wrote in a decision granting the plaintiff's attorneys' request for sanctions that the defendant's counsel, Michael Fourte's law offices, not only submitted AI-hallucinated citations and quotations in the summary judgment brief that led to the filing of the plaintiff's motion for sanctions, but also included "multiple new AI-hallucinated citations and quotations" in the process of opposing the motion.
"In other words," the judge wrote, "counsel relied upon unvetted AI -- in his telling, via inadequately supervised colleagues -- to defend his use of unvetted AI."
The case itself centers on a dispute between family members and a defaulted loan. The details of the case involve a fairly run-of-the-mill domestic money beef, but Fourte's office allegedly using AI that generated fake citations, and then inserting nonexistent citations into the opposition brief, has become the bigger story.
probably needed AI to generate an explanation (Score:1)
Re: (Score:2)
No one will know if you actually know what 'irony' means.
Re: (Score:3)
Disbar the fucker (Score:5, Insightful)
I am sure he still charges the usual exorbitant rates even when he fails to supervise the AI. Ordinarily, that is called fraud. I am sure the legal mafia has made themselves some laws around that.
Re: (Score:2)
Flood the zone with shit, the gold standard of modern life. It only matters that the rewards are greater than the consequences.
Re:Disbar the fucker (Score:4, Insightful)
Pretty sure this opens him up to a legal malpractice suit. Probably more lucrative than whatever the debt was.
Disbarment and blocking his firm from the court (Score:2)
These should be legal ethics cases with a high likelihood of disbarment for the attorney and blocking of his law firm, and its partners from appearing before the NY supreme court and other NY state courts, working on, advising on, filing documents with the NY state courts for 5 years including disclosing the name of any of those attorneys on all documents.
Ballsy move (Score:5, Insightful)
This guy has done the improbable and somehow managed to lower my opinion of lawyers. It was already a low bar, but this man is the new limbo champion. The only saving grace here is that this is a civil matter and no one will be going to jail for this man's failings. Perhaps he should though, if for no other reason that to dissuade anyone else from acting as foolishly.
Re: (Score:1)
Come on, I see great promise in him. He should send his candidature for a job at the White House. Or maybe run for congress, republicans love people like him.
Wait, one question though, did he admit to using AI. If he did, then forget I said anything. Remember kids, take it from Trump, never admit to any wrongdoing. Accuse your accuser of what you did. So if you get caught using fake cases in court, deny they are fake, accuse them of having deleted the cases just to put him in jail, and that they used fake c
Re: (Score:2)
You give him too much credit. It's really just a epically stupid move. "Ballsy" implies that the guy *knew* he was taking a risk.
Re: (Score:2)
You give him too much credit.
I totally agree.
It's really just a epically stupid move.
I totally agree again.
"Ballsy" implies that the guy *knew* he was taking a risk.
He knew. He just didn't care, as judges have stupidly been VERY light-handed about attorney AI fraud. If judges did the correct thing, and heavily sanctioned attorneys who openly defraud courts via LLM's, then the problem would resolve itself rather quickly.
Re: It's going to be interesting to see what happe (Score:2)
Re: (Score:2)
It's only a matter of time before AI can replace most common high profit lawyer work.
Possibly... but it won't be with LLMs. They're inherently unreliable and absurdly expensive. There's a reason why 95% of AI projects fail, and I consider that number incredibly optimistic.
Re: (Score:2)
This is from 2022.
https://www.superlegal.ai/blog... [superlegal.ai]
AI vs Lawyers – the Results
When it comes to lawyers vs AI – it wouldn’t be a huge surprise if we told you that the AI was faster, but would you believe us if we told you that the AI was also more accurate?
Results:
On average, the lawyers took 92 minutes to finish reviewing the contracts compared to the AI taking 26 seconds.
The AI finished the test with an average accuracy rating of 9
Re: (Score:2)
Contract review and document review are two places where AI could be pretty good.
On the other hand I'd be a bit skeptical about using it for contract creation without a skilled lawyer knowing how to guild the AI to make sure the important points are covered. Much less creating legal filings, where hallucinating cases is getting people disbarred.
Re: (Score:3)
You have absolutely no clue what is high profit legal work.
Partners at AM100 law firms doing Mergers and Acquisitions typically bill over $1,500 an hour and can bill as much as $3,000 an hour. They can have a dozen partners bill 100 hours a week for months at those rates. And you think divorce is high profit?
This is why nobody takes AI fan boys who think AI will take over the legal profession seriously.
You are talking about the low margin stuff that people wish they weren't doing, and calling it high margin
Re: (Score:2)
(Hallucinated) Turtles All The Way Down (Score:2)
https://en.wikipedia.org/wiki/... [wikipedia.org]
Until the sanctions for the firms and lawyers are (Score:5, Insightful)
Re: Until the sanctions for the firms and lawyers (Score:2)
This.
How is he not immediately disbarred?
AI-hallucinated citations :o (Score:3)
Re: (Score:3)
It's a combination of training data and rewards. Chatbots are trained to never admit that they don't know, and to always be willing to be convinced that the person talking to them is correct. This makes them more popular, and enchances engagement, but at the cost of accuracy.
Re: (Score:2)
Chatbots are trained to never admit that they don't know, and to always be willing to be convinced that the person talking to them is correct.
No, that's exactly not what chatbots are doing. Chatbots have no concept of right and wrong. Chatbots know that given the frequency of words already in the conversation and their probalistic neighborhood to elements in their body of data, which words are most probable to come next. And if there is not enough data fitting their current state, they randomly add words, because no possible next word presents them with a high probability.
Re: (Score:2)
Chatbots know that given the frequency of words already in the conversation and their probalistic neighborhood to elements in their body of data, which words are most probable to come next. And if there is not enough data fitting their current state, they randomly add words, because no possible next word presents them with a high probability.
The old n-gram models were kind of like this. Modern systems use a neural model and have the ability to generalize. In other words they actually do and learn shit. For example solving simple ciphers, base64 decoding, language translation, ICL..etc.
There is some basic statistics /w random probability at the end of the round after a model is run through in selecting a token yet this is a footnote that is in no way representative of the internal processing and communication that takes place during execution
Re: (Score:2)
It's a combination of training data and rewards. Chatbots are trained to never admit that they don't know
They are not trained to do this. They have no meta-cognition and don't know what they know in the first place.
I'll wait for the final word (Score:3)
Non paywalled article ... (Score:4, Informative)
Here is a non-paywalled article that even has a summary at the top:
NY judge sanctions lawyer for fake AI citations [nydailyrecord.com].
Grow a pair FFS (Score:2)
Why not just use AI... (Score:2)
to tell the judge why the stupid human used AI.
As a matter of fact, you don't need the stupid lawyer at all!
TDOJ has a position for him (Score:2)
The Trump Department of Justice is looking for men of his caliber and outstanding integrity. Remember children, no one except Trump and his cronies are above the law.
Volokh Conspiracy covered this too (Score:5, Informative)
https://reason.com/volokh/2025... [reason.com]
It's a legal blog with a legal perspective.
Re: (Score:2)
"And, your Honor, when I say I told the staff, I take full responsibility. It's my staff, so I told myself to get rid of it and I did not get rid of it"
Don't you hate it when you tell yourself to do something, and you don't listen? Ugh. If you can't depend on yourself, on whom can you depend?
Allow me to introduce myself your honor (Score:2)
Sloppy McSlopperson, Attorney-At-Slop
"Vibe law" (Score:2)
hmm, maybe backwards (Score:2)
Maybe AI is using lawyer to explain what it did?