Every day this week so far Law360 has reported on a new AI-related sanctions order, and I’m sure these are just the ones Law360 made space for. Pretty soon they’ll need a separate newsletter just for that topic (and if they do, I will subscribe to it). I too do not have space (well, time) to write about all of these, so there needs to be some distinguishing characteristic for a story to qualify. “Dummy relies on generative-AI output” is no longer sufficiently newsworthy.

Today’s sufficiently-newsworthy example involves a decision by the Supreme Court of Alabama, which not only sanctioned the AI-using attorney but dismissed his clients’ appeal because it found the briefs were “grossly deficient.” Ibach v. Stewart, No. SC-2025-0106 (Ala. Apr. 24, 2026). The facts of the case are unimportant to us, until we reach Section II, entitled “[Counsel’s] Extensive Use of Nonexistent or Misquoted Authorities.” How extensive? “Astoundingly” so, it turns out.

This was first identified, as usual, by opposing counsel. See, in the U.S. we have an “adversary” system, meaning one is normally opposed by an adversary who might do adversarial things, like actually look for the cases you cited and tell on you if you made them up. I’m sure all of you know that, but it’s starting to seem like a lot of people don’t. Anyway, that’s what happened here. In fact, according to the opinion, the response brief spent twelve pages talking about how many of the moving party’s cases didn’t exist. That seems like too many, but it’s hard to fault those lawyers for beating up on someone who fabricated an “astounding number” of authorities.

And that was just in the opening brief. In the reply—which, remember, was supposed to be replying to a brief that spent twelve pages talking about how many of the moving party’s cases didn’t exist—the offending attorney addressed the problems only in a single footnote. Here’s the footnote:

[Plaintiffs] acknowledge and regret that their opening brief misquoted two secondary sources …. The error arose from counsel’s first use of an AI research tool that summarized commentary not readily available through standard legal databases. The tool misattributed quotes. Counsel accepts full responsibility for relying on those summaries without independently verifying the original texts. The mistake will not recur. But the underlying legal principle … is both correct and well-settled. See Ex parte Seabol, 782 So. 2d 212, 216-17 (Ala. 2000); Ex parte United Serv. Auto. Ass’n, 78 So. 3d 979, 983-84 (Ala. 2011); Hughes v. Glover, 157 So. 2d 299, 302 (Ala. 1963); Franciscan Sisters Health Care Corp. v. Dean, 448 N.E.2d 872, 876 (Ill. 1983); Restatement (Second) of Trusts § 219(1) & cmt. a.

Have you guessed the problem? I put some of the words in bold to help you spot it. Correct! Counsel’s promise, “the mistake will not recur,” was fulfilled for one sentence.

I could also point out (the court certainly did) that the problem went far beyond “two secondary sources,” and also that the rest of the brief was also full of bogus citations. In fact, the next fourteen pages of the court’s opinion was devoted exclusively to those bogus citations.

At the show-cause hearing, the attorney finally admitted what he had done—or more accurately, not done—and showed remorse, and a week later he wrote the other side a check for $17,200. The court still alerted the state bar, and denied the plaintiffs’ motion to file supplemental briefs that had real legal authorities in them. Appeal dismissed.

Two justices dissented in part, mainly from the decision to dismiss the appeal, which punishes the clients for something that was the lawyer’s responsibility. (He will be hearing from their new lawyer soon.) But they all agreed that they’re just about fed up with this nonsense.

This is not to say, as even the majority pointed out, that it is necessarily wrong or unethical to use generative AI for … whatever you’re using it for. But it is both wrong and unethical to rely on its output without confirming the results yourself. It is designed to generate things that look like humans wrote them, not necessarily to generate right answers. That’s the human’s job.

Similarly, an AI could probably generate what looks like a check for $17,200, but for now it’s still the human’s job to pay the money. I would keep that in mind too.

       

Related Stories

• Judge Prepares Slide Deck of Lawyer's Mistakes
• In the Ancient Libraries of Ashurbanipal
• Please Do Your Best Not to Appear in the “AI Hallucination Database”

 

I do like a good hyphenated adjective, and this case definitely called for one.

It’s a little surprising (though maybe it shouldn’t be) that it took more than a year for this to work its way through the justice system. But it did. As you may recall, what authorities called “Operation Bear Claw” ended in November 2024 with the arrest of four (4) suspects and the confiscation of at least one (1) bear suit. See “Insurer Thinks Bear Shown in Video Trashing Car Looks a Lot Like a Guy in a Bear Suit” (Nov. 14, 2024). Frankly, that headline should answer a lot of the questions you might have about what happened, but the 2024 article has further details plus a large full-length image of the aforementioned bear suit. But to summarize, the culprits clawed up, or conspired to claw up, a Rolls-Royce and two other luxury vehicles, which they then told insurers had been damaged by bears.

Apparently expecting the insurers to doubt this very plausible and even commonplace explanation for vehicle damage, they helpfully submitted alleged security-cam videos in support of their claims. I should stress that in all three videos, the alleged bear was inside the vehicle. While bears apparently can learn to open car doors, investigators thought it was unlikely that what appeared to be the same bear would climb into three different luxury vehicles at about the same time and place (as the videos showed when compared), and yet not test-drive even one. Clawing up the outside of the cars would have been more plausible, but (1) that probably would have been easier and cheaper to fix, resulting in a lower payout, and (2) it would have been much easier to tell that these weren’t real bears.

Even so, I’m not sure it takes an expert to determine that the beast(s) seen thrashing around in the cars weren’t bears. But it turns out that the State of California has bear experts who are able to tell the difference between “bear” and “guy in a bear suit” even under these circumstances.

According to Insurance Journal (and let me pause here to thank my source for reading Insurance Journal so I don’t have to), the bogus claims amounted to $141,839. On April 16, the California Department of Insurance announced that three of the four defendants had pleaded no contest to felony insurance fraud. Each was sentenced to 180 days in jail, plus probation and restitution. The fourth is due for a preliminary hearing in September. No reason for the delay in his case was given, which leaves me free to speculate that he took a backup bear suit and fled into the Sierras. But I think there are still some real bears left up there, and if so they may have encouraged him to go back and face justice.

“Insurance fraud is a serious crime that drives up costs for consumers,” the department said in a statement, adding that “no scheme is too outrageous for us to investigate.” That last part sounds good but I’m not really sure what it means. Aren’t outrageous schemes the most likely ones to be investigated? (Assuming the White House isn’t involved, I mean.) Are there insurance-fraud schemes so outrageous that other departments of insurance just throw up their hands and won’t look into them? “No, no–too outrageous!” Please advise.

See also, e.g., “Costumed Bear Harasser Wanted by Authorities, Evolution” (Aug. 17, 2015) (wondering why someone would dress up like a bear to harass a real bear); “How to Get Back at Your Ex: First, Kill a Bear” (Oct. 3, 2011) (discussing a bear-suit plan that failed at the first step).

       

Related Stories

• Cat Blamed for Crime
• Alleged Irish Orphan Convicted of Being American Rapist
• Defendant Recruited Seven Actors to “Testify” in Fraud Trial

 

Law360 reports that the Fourth Circuit has affirmed the conviction of a man who illegally accessed and leaked the private health information of former Justice Ruth Bader Ginsburg. In United States v. Russell, the court held it was not an abuse of discretion to deny Russell’s motion to exclude statements he made to federal agents during an interview.

Those statements included Russell’s attempt to blame his cat for the incriminating search term, “gins,” which Russell suggested his cat had typed by running across his keyboard.

Before her death in 2020, Justice Ginsburg was being treated at George Washington University Hospital. In January 2019, someone posted information about her treatment on 4Chan, where, the opinion says, users were “promot[ing] a conspiracy theory that Justice Ginsburg had [already] died and prominent Democrats were covering up her death.” (Presumably the motive would have been to prevent then-President Bozo Waffles from replacing Ginsburg. I would have supported that, but the idea that “prominent Democrats” would have been competent enough to pull it off is almost as comical as the cat-blaming.) The information then spread to Twitter.

Wrongfully obtaining and disclosing such information is a federal crime (well, two crimes), and an investigation followed. Not surprisingly, this focused on hospital employees and contractors who had access to the hospital’s computer system. Should a hospital employee or contractor with access to the system have foreseen this? Probably. But at least two did not.

Hospital officials looked at the search logs to see if anyone not involved with Ginsburg’s treatment had searched for a name starting with “ginsb” during the relevant period. They found that an ER employee had searched for “ginsbur,” but decided it was unlikely he had been the leaker (he was fired, but not prosecuted). So they kept looking.

This led them to Russell, a contractor working as a transplant coordinator. On January 7, 2019, three searches were run using Russell’s credentials: “gins,” “barker,” and “ginston.” There was a patient named “Barker,” but no “Gins” or “Ginston.” But there was, of course, a patient named “Ginsburg.” This was enough for a visit from federal agents. Also a visit from Russell’s boss, who decided she wanted to sit in on the interview. So she was present for the lies Russell was about to tell.

Russell admitted that whoever ran the searches had been logged in as him. He also admitted he had searched for “barker,” which he might have had a legitimate reason to do. But he denied searching for “gins” or “ginston.” Hm. So who might have run those searches while logged in as him, immediately before and after a search he admitted he did run?

He didn’t know, he told the agents, but he had a couple of theories.

The first was that “a coworker may have used his login information” to run the other two searches. He was being framed! Probably not, though, because the agents knew the three searches had been run within a matter of seconds. (Prosecutors argued Russell realized, correctly but too late, that “gins” might look suspicious, so he followed that with a fake but plausible search and then a search for a made-up name also starting with “gins,” hoping to cover his tracks.)

But the second theory is why we’re here: Russell suggested that “potentially his cat had run across the keyboard and typed in those letters.”

It is, of course, statistically true that if an infinite number of cats ran across an infinite number of keyboards, sooner or later one of them would type the name of a U.S. Supreme Court justice. And—to my surprise—it is possible that even a finite number of cats and keyboards might accomplish this during the estimated lifespan of the universe. See Stephen Woodcock and Jay Falletta, “A numerical evaluation of the Finite Monkeys Theorem,” 9 Franklin Open Journal (Oct. 30, 2024).

Woodstock and Falletta found that, even if 200,000 monkeys had 200,000 keyboards, it would be effectively impossible for them to produce the complete works of Shakespeare (the usual goal in this thought experiment) before the universe comes to an end. Id.; see also “Monkeys will never type Shakespeare, study finds,” BBC News (Oct. 31, 2024). But towards the other end of the scale, there was a non-zero chance—about 5%, in fact—that just one monkey would produce one particular word (“bananas”) during its estimated lifespan. (I know—for some reason they start talking about chimps, which aren’t monkeys. Email them, not me. We’ve been over that. See “Clarification: Monkey v. Ape,” Lowering the Bar (Oct. 20, 2008)).

So I conclude it wasn’t impossible that during its lifetime, Russell’s cat could have randomly typed the name of a U.S. Supreme Court justice, even Justice Ginsburg in particular. A fortiori, it could have typed the shorter terms “gins” or “Ginston” during that time. Sadly, Russell didn’t have the benefit of this research in 2019. But it wouldn’t have helped anyway. Even accepting the above, the probability the cat would have typed both terms within a mere matter of seconds (pausing to let Russell search for “Barker”) would surely be effectively zero.

As you’ve probably guessed, this wasn’t the only problem with Russell’s story. When agents asked to see his work computer, he said it had been stolen. (Also not impossible!) But he did turned over a hard drive from his home computer, and even though he had tried to wipe it, the government recovered enough information to indict him. (Destroying evidence became his third charge.)

What, you want (another) legal issue? Fine. The main issue on appeal was whether the district court should have excluded what Russell said during the interview and thus also the hard drive. Russell argued the statements were “involuntary,” not because he was beaten or anything but because his company’s CEO was there. But the agents didn’t invite her—she just decided to sit in—and the court held the circumstances didn’t establish coercion. Russell volunteered his very unlikely story.

Of course, given enough judges and enough opportunities, it’s not impossible Russell might convince one with arguments like these. It’s just not likely to happen during the estimated lifespan of the universe.

       

Related Stories

• Sentences Handed Down in Bear-Suit-Insurance-Fraud Case
• Alleged Irish Orphan Convicted of Being American Rapist
• Defendant Recruited Seven Actors to “Testify” in Fraud Trial