Another day, another fictitious case citation. Blink, and you’ll miss the most recent comedy of AI-created errors.
Since the release of ChatGPT in late 2022, the frequency of court submissions riddled with AI-hallucinated gibberish has increased exponentially. Now, more than three years later, it seems that not a week goes by without a headline about yet another case in which a lawyer has submitted briefs to the court full of AI-hallucinated gibberish.
One of the standout features of many of these cases is that the attorneys double down, rather than admitting the error of their ways. Sometimes, they even submit responsive papers in defense of their actions that include hallucinations.
Of course, it’s one thing to read about these shenanigans, but seeing the audacity in action during an appellate argument? Priceless. And appalling. In equal proportions.
I present to you, attorney for the appellant in Deutsche Bank National Trust Company v. Jean LeTennier, CV-23-0713 (2026), whose absolute chutzpah was on full display — captured on video on October 16, 2025, during an oral argument before the New York Appellate Division, Third Department.
About one-third of the way into the video, at 6:30, he faced a very hot bench. The judges were collectively, and understandably, piqued by both the number of hallucinations contained in his submissions and the fact that he seemed to be entirely unbothered by his own fictions.
When asked about his response to the allegations that he’d used AI, he appeared to be oblivious to the very admissions he’d included in his responsive papers. Rather than answering their pointed questions about the errors, he deflected and tried to avoid the topic entirely*:
The Court: The citations used in some of the cases are not real citations (to) real cases.
Counsel: (T)hat was never asserted to me.
Humans in the Loop: The People Powering Trusted Legal AI
As the use of artificial intelligence permeates legal practice, a critical question confronts every legal professional who uses these tools: Can I trust this?
The Court: You acknowledged it in your reply brief that some of the cases were not accurate.
Counsel: The issue that I believe is really important is the fact that fraud was…
The Court: It’s important to us, so I guess you’re going to have to deal with what’s important to us.
The judges were not dissuaded by his obfuscation. They persisted in their questioning, forcing him to concede that there were fake cases cited in his briefs. Undeterred, he then informed the panel that the AI issue was immaterial, apparently deciding that patronizing the bench was a winning tactical move:
The Court: Did you, as an attorney, write a brief and submit it to this court?
Counsel: Yes,
Court: Okay, so you own it, right?
Counsel: Yes.
Court: Okay, in that brief, we’re telling you, and you are aware … not only were you made aware by the court, but by your adversary, that there are citations that are not real cases.
Counsel: That’s not germane to the fact that the SEC is telling this court …
Once that approach proved ineffective, he switched gears once again. When directly questioned about his AI use, he played coy, acting like a petulant adolescent who’d been caught drinking. Finding that tactic to be futile, he once again reverted to his previously unsuccessful strategy of mansplaining to the bench:
The Court: So, I guess I’m asking you, did you use AI to do the brief, and are these hallucinated cases, or did you miscite cases? Because you didn’t give us any corrections …
Counsel: AI is a tool that I think all of us use these days.
The Court: So that would be a “yes, I used AI.”
Counsel: Well, not exactly. I mean, yeah, I used AI.
The Court: Okay. You’ve got to check AI, right?
Counsel: I do.
The Court: Well, evidently not too well, right?
Counsel: It seems like we’re not able to focus on the issue that they brought it …
When it became apparent that condescension wasn’t working, he retreated to a defense of statistical probability. The exchange that followed captures the surreal moment that he attempted to treat a ‘mostly accurate’ brief as a job well done:
The Court: Because we’re on this side of the of the bench … that’s why we’re asking you about the citations in your brief that you provided to us and the response when it was pointed out that these are AI citations.
Counsel: I believe that the citations that I used were accurate, like 90% were accurate, some of them which really aren’t necessarily germane to the issues at hand …
The Court: Okay, your time is up! Thank you.
Needless to say, the court was unimpressed with his assertion that a 90% accuracy rate was a passing grade for the truth, dismissing his argument in its written decision, issued in January: “(D)uring oral argument defense counsel estimated that 90% of the citations he used were accurate, which, even if it were true, is simply unacceptable by any measure of candor to any court.”
In the court’s eyes, a brief that is only 10% imaginary is still 100% problematic, especially when that small slice of fiction accounted for “at least 23 fabricated legal authorities across five filings … (and misrepresenting) the holdings of several real cases as being dispositive in his favor — when they were not.”
His stubborn resistance to reality was rewarded with sanctions in the amount of $5,000 due to his refusal to take accountability for his actions: “(H)is reliance on fabricated legal authorities grew more prolific as this appeal proceeded … Rather than taking remedial measures or expressing remorse, defense counsel essentially doubled down during oral argument on his reliance of fake legal authorities as not ‘germane’ to the appeal.”
Importantly, the court acknowledged that AI does have a place in litigation, as long as attorneys and staff are sufficiently trained and carefully check their work for accuracy before submitting it to the court: “As with the work from a paralegal, intern or another attorney, the use of GenAI in no way abrogates an attorney’s or litigant’s obligation to fact check and cite check every document filed with a court. To do otherwise may be sanctionable …”
It’s a bold new era for the legal profession: one where ‘mostly accurate’ is a tactical hill to die on and 23 imaginary cases are just ‘minor’ details. As it turns out, that final 10% is the difference between a winning argument and a $5,000 audacity tax.
* The court transcript excerpts have been lightly edited for readability and flow.
Nicole Black is a Rochester, New York attorney and Principal Legal Insight Strategist at 8am, the team behind 8am MyCase, LawPay, CasePeer, and DocketWise. She’s been blogging since 2005, has written a weekly column for the Daily Record since 2007, is the author of Cloud Computing for Lawyers, co-authors Social Media for Lawyers: the Next Frontier, and co-authors Criminal Law in New York. She’s easily distracted by the potential of bright and shiny tech gadgets, along with good food and wine. You can follow her on Twitter at @nikiblack and she can be reached at [email protected].