Lowering the Bar Lowering the Bar 2026-06-19T23:04:12Z https://www.loweringthebar.net/feed/atom WordPress https://www.loweringthebar.net/wp-content/uploads/2018/06/cropped-LTB-512x512-1-32x32.jpg https://www.loweringthebar.net/2026/06/assorted-stupidity-174.html Kevin https://www.loweringthebar.net <![CDATA[Assorted Stupidity #174]]> https://www.loweringthebar.net/?p=47256 2026-06-19T23:04:12Z 2026-06-19T23:03:46Z In this edition: a FOIA request; another psychic fails to predict her own future; important disclosure requirements; etc.
       

Related Stories

 
]]>

  • CORRESPONDENCE/MARK FOR FOLLOW-UP: Dear Kent Police Public Disclosure Team: I still have not received a response to my Freedom of Information Act request of 28 May 2026 asking you to provide “the number of people aged nine or below who have been reported for criminal offences” in your county. Please provide this important data at your earliest convenience. Yours very sincerely, Kevin Underhill
  • “We were like, oh, OK, this is happening, like, now,” Samantha Randazzo’s lawyer told the New York Times. What was happening was, like, his client was totally giving birth in a courtroom, after her water broke while she was waiting to be arraigned on drug charges. Explaining why she was cuffed (at least at first), the NYPD insisted it had no idea she was pregnant because she was “wearing baggy clothes” and never mentioned a pregnancy. From TV shows I gather that suspects are usually searched incident to arrest, but police would never lie about something like this. “We are delighted both mother and baby are well,” said a spokesperson for the Office of Court Administration, in that spokesperson voice.
  • According to Renea Gamble’s daughter, her mom intended to go to a No Kings protest in October 2025 dressed as a sea turtle and holding a sign that said “I love the Gulf of Mexico.” But that costume wasn’t delivered on time, which is why her sign said “No Dick Tator” and she was dressed as a seven-foot-tall representation of its second word. Because this is America, she was thrown to the ground, arrested, and charged with a crime. But also because this is America (for now), a packed courtroom broke into applause after she was acquitted in April. The brave officer who tackled and subdued the 62-year-old testified under oath that he did so because he thought her costume posed a “safety risk.” Seems unlikely, but police would never lie about something like this.
  • If you’re selling your home, do you have a duty to tell potential buyers that your neighbors are awful? In this lawsuit, a buyer who was later shot by such a neighbor alleges that the failure to disclose his awfulness violated a Minnesota law requiring disclosure of anything that would “significantly affect an ordinary buyer’s use and enjoyment of the property.” The report says the seller had filed a restraining order against the neighbor in 2025, but doesn’t say what prompted that.
  • Don’t forget that in some states, you may have a duty to disclose whether your house might be haunted, or at least whether anyone has died on the property within a certain number of years. See, e.g., “House Advertised as ‘Not Haunted‘” (Oct. 31, 2024). Not that the legislatures believe ghosts are real, mind you, they just think people should know they may be buying, as Connecticut used to call it, “psychologically impacted property.”
  • If a speed camera with a +/- 1 mph margin of error clocks a driver at 61 mph in a 50-mph zone, is the driver guilty of exceeding the limit by “11 to 15 mph”? According to the District of Columbia, yes, because a regulation “deems” a certified camera’s measurements to be perfectly accurate. According to the D.C. Court of Appeals, no. Given the margin of error, it held, the District proved only that he was traveling somewhere between 60 and 62 mph, which is not the same as “at least 61.” Frankly, this opinion could have been less than 23 pages long, but it comes out right. Congratulations to the defendant; not so much the D.C. Attorney General’s office, which seems to have paid five lawyers to fail to enforce a $50 fine.

       

Related Stories

 
]]> https://www.loweringthebar.net/2026/06/its-finally-happened-both-sides-ai.html Kevin https://www.loweringthebar.net <![CDATA[It Was Only a Matter of Time: *Both* Sides’ Lawyers Sanctioned for Using AI]]> https://www.loweringthebar.net/?p=47281 2026-06-09T21:18:15Z 2026-06-09T21:17:11Z None of the four lawyers involved will be seeing the inside of this courtroom again for a while.]]>

On several occasions here we’ve discussed the remarkably stupid tradition of dueling, I guess most recently in this 2024 item about a probably-not-serious proposal to embed that practice in the rules of the Missouri Senate. Lawyers were never immune to this stupidity, which is why, for example, new members of the Kentucky Bar were required to swear an oath that they had never “fought a duel with deadly weapons” or aided or assisted in such a thing. And, in fact, they still are. Today I was interested in finding out whether there had ever been a duel in which both idiots died. I thought I had found three examples, but it turned out those were all bullshit.

That’s what I get for using AI to run the search.

It seems like only yesterday, but in fact was four days ago, that I mentioned the latest fiasco in which lawyers got sanctioned because they had relied on generative AI without checking its output. See “Two More Lawyers Pay for Relying on Artificial ‘Intelligence’” (June 5, 2026). That case was worse than most because the lawyers also weren’t totally honest with the court, although this is something that happens surprisingly often in AI-related sanction cases. According to the AI Hallucination Database, there have been 1,598 cases to date, worldwide, in which someone got in trouble for relying on a generative AI tool that turned out to have just made stuff up.

Of course I haven’t read all of those, but I suspect this might be the first time that all the lawyers on both sides were kicked off the case for doing that.

By “this” I mean Monday’s opinion in Withers v. City of Aberdeen, decided by the federal court for the Northern District of Mississippi. The court had issued an order “directing all counsel of record in this case to show cause as to why the Court should not impose sanctions against them pursuant to [various rules] for filing legal memoranda containing hallucinated case citations.” (Emphasis added.) Each party was represented by two lawyers, one out-of-state, one local. The two out-of-state lawyers admitted they had used generative AI (one for research, the other to draft the brief), and also that neither one bothered to verify what the AI had given them. The two local counsel said they were “unaware of their respective co-counsel’s use of AI,” but admitted that they committed essentially the same error by not reviewing what the other humans had given them. They all then suffered through a show-cause hearing that was surely quite embarrassing.

After discussing the facts, the opinion says this: “Following the hearing, the Court stayed this case and cancelled the scheduled trial, which was previously set to occur on March 23, 2026.” Well, that was bad news for someone, and it was probably no great surprise that it turned out to be bad news for all four.

“This Court is yet again burdened with addressing AI hallucinations in court filings,” the opinion says, citing a number of prior cases. “It has previously acknowledged that AI is a powerful tool that when used prudently, provides immense benefits.” (Emphasis in original, scoffing at the word ‘immense’ added by me.) “However,” the court continues, while generative technology can produce words, it cannot attach sincerity, truth, or responsibility to what it writes.” (Citation omitted.) That’s a human’s job.

Again, none of this is new—which is exactly why the court wasn’t buying one lawyer’s claim that “she was shocked when the Court issued the show cause order,” essentially taking the position that “she was unaware that AI could produce hallucinated cases and explained that she did not even know what a hallucinated case was.” Nope. “The Court finds that explanation to be insufficient and incredulous.” (I think technically the court would be “incredulous” and the explanation “incredible,” but please don’t tell the court I said that. It’s already mad enough.) The increduli-ble-ness was primarily because this has happened and been publicized so many times now that one can expect a reasonable lawyer to know about it—meaning, this excuse might have worked earlier but its time has run out.

The opinion also suggests that deflecting by arguing there are real cases that support the proposition for which you offered the fake ones is not going to work anymore, either (if it ever did). Another lawyer offered that one to the court and it only made it angrier.

Each lawyer gets his or her own section in the opinion, which is not something you want to have happen, unless maybe you saved the judge’s life at some point. To sum it up, the court found varying levels of responsibility, but the results weren’t that different. In particular, all four attorneys were booted from the case. The out-of-state attorneys had their pro hac vice admissions revoked and cannot show their faces in the Northern District of Mississippi (well, its courthouse) for two years. The in-state attorneys were disqualified from appearing further in this case. All four were fined in varying amounts, and the court referred all four to the relevant state bar associations. So it seems fair to say that using generative AI (imprudently) did not produce immense benefits in this case.

I should point out that all this resulted from “only” six fake cases, four on one side and two on the other. I don’t know what percentage that was, but I’m pretty sure it doesn’t matter unless the percentage in your brief is zero.

]]> https://www.loweringthebar.net/2026/06/two-more-lawyers-pay-for-ai-reliance.html Kevin https://www.loweringthebar.net <![CDATA[Two More Lawyers Pay for Relying on Artificial “Intelligence”]]> https://www.loweringthebar.net/?p=47277 2026-06-05T23:06:51Z 2026-06-05T23:06:47Z Is it worth it? (That's a rhetorical question.)]]>

As we’ve discussed here before and, unfortunately, as people will probably have to keep discussing from now until the Sun grows old and expands to consume the Earth, if “people” survive that long, artificial “intelligence” is perhaps not all it’s cracked up to be. Generative AI has its uses, you say? Okay, then go use it for those uses. But don’t rely on what it tells you in any situation that really matters, or you are likely to regret having done so.

Also, if you get caught relying on it, my advice is not to lie about this to the court.

It’s too late for the lawyers who got sanctioned by the Ninth Circuit for this on June 3, as detailed in the opinion below. Well, I shouldn’t suggest that they lied. Let’s say they “violated their duty of candor” when they were less than “transparent” when explaining the source of the many obvious errors in their brief, and thus “knowingly or recklessly made false statements” to the court. Because that’s how the court put it. It also noted other problems with the logic of some of their explanations, demonstrating, it said, that the lawyers’ “post-hoc rationalizations are thus not forthright.” And man, just like your parents always told you, you don’t want to be in a situation where your post-hoc rationalizations may be seen as not forthright. That is not good.

I’m no Biblical scholar, but I think that was in the first draft of the Ten Commandments, wasn’t it? “Thou shalt not be less than forthright in thy post-hoc rationalizations”? Good thing somebody did some editing before those were released to the public. (Sorry, I meant “Somebody.”)

This case is not so different from the hundreds of cases to date in which somebody (small S) has been punished—not just embarrassed, but punished—for relying on generative artificial intelligence instead of their own. Just using these things is not itself sanctionable, as many courts have explained. But there seems to be something about these tools that makes people assume they can rely on the output without double-checking it. Don’t do that.

And again, don’t be less that forthright if you get caught. As the Ninth Circuit stressed, “[t]here is no upside” to that.

We stress that when an attorney learns of any error in a filing—including generative AI hallucinations—he should immediately alert the court and opposing counsel of the error and disclose its source. There is no upside to denying the use of generative AI or to passing off an AI hallucination as an innocent typographical error. The other rules discussed in this order—competence, diligence, meritorious arguments, citations to authority, attestations to accuracy—do not turn on the source of the error.

Sanctions: ordered to pay $2500 each, forbidden to practice before the Ninth Circuit for six months, forced to give a copy of this embarrassing order to pretty much everybody except their grandparents, required to include an “AI disclosure” under penalty of perjury in all filings for the next two years, and referred to the state bar for possible further punishment.

There is no upside!

]]> https://www.loweringthebar.net/2026/05/good-reason-to-kill-79-disputed-seating.html Kevin https://www.loweringthebar.net <![CDATA[Good Reason to Kill #79: Disputed Seating at Kindergarten Graduation]]> https://www.loweringthebar.net/?p=47264 2026-05-29T06:04:55Z 2026-05-29T06:02:38Z "A kindergarten graduation turned violent today," but only for the parents.]]>

“A kindergarten graduation turned violent today” were the first words of this report from WTVG about a May 21 incident at Queen of Apostles School in Toledo, Ohio. At least a dozen people, all chronologically adult, ended up in a brawl after being unable to resolve a dispute over seating for the ceremony. One woman was arrested and charged with “felonious assault” for allegedly grabbing another woman by the hair and banging her head into a chair. The loser of that exchange needed stitches, according to the report.

Thankfully, no children were harmed, and no serious injuries were reported and certainly no deaths. As we’ve discussed before, the name of this category is to be understood sarcastically, not literally. If you wish further elaboration, you can find it at the beginning of prior installments. See, e.g., “Good Reason to Kill #78: Ate the Last Hot Pocket” (May 30, 2023); “Good Reason to Kill #77: Summoned Bigfoot to Kill You First” (July 13, 2022).

Okay, that last one did involve a murder, but that was an exception to the general rule, because Bigfoot.

No one was killed at the kindergarten ceremony, that’s the point. But there was a full-on brawl between various nominally adult family members, with punches thrown, as you can see on WTVG’s site because of course someone recorded it on their phone. When introducing that clip, the anchorperson felt obligated to say that it was “disturbing,” by which I assume he meant “hilarious.”

It doesn’t show how the dispute arose, or how a dispute could have arisen about seating at a kindergarten graduation ceremony. According to a witness, one family “started to grab their own chairs and kind of like make their own space,” apparently in front of others who then couldn’t see well. The video seems to show there is plenty of space in the room to which one could relocate, and that the ceiling is high enough that standing up would also have been an option. Doing that would put one’s eyeballs above the back of a seated person’s head, thus ensuring that photons bouncing off the children could reach those eyeballs and thereby the attached brain for analysis. But it seems no one thought of that.

According to the witness, he was involved in some of the back-and-forth before things turned violent. At some point, he said, a “whole family in the first two rows stood up—five guys, five girls—they all just stand up,” and before he knew it he had been “sucker-punched.” This knocked him to the ground, whereupon “probably four or five other guys were on top of me, trampling me, punching me, kicking me in the head.” The video generally confirms this, but the man doesn’t seem to have been in any real danger. Let’s just say most of the blows look like they could have been delivered by the children, not just witnessed by them.

Nor is a felonious assault immediately apparent from watching the video. Of course, that depends on what “felonious assault” means in Ohio. At common law, and under some statutes, an “assault” is an intentional act that creates a reasonable fear of imminent harm, whether contact is made or not. Making contact is “battery,” and of course these usually go together—except in the case of a sucker punch, which would be battery unpreceded by assault. But never mind that, because Ohio statutes don’t distinguish between assault and battery. They just define different flavors of assault. In rough order of badness, these are: negligent, regular, aggravated, and felonious.

  • Regular “assault” means knowingly causing or attempting to cause physical harm to another, or recklessly causing it. Generally a first-degree misdemeanor (i.e., the worst kind), but circumstances could make it a fifth-, fourth-, or even third-degree felony.
  • “Aggravated assault” is more complicated. This means knowingly causing serious physical harm, or causing or attempting to cause any physical harm by means of a deadly weapon or dangerous ordnance, “while under the influence of sudden passion or in a sudden fit of rage” brought on by “serious provocation occasioned by the victim that is reasonably sufficient to incite the person into using deadly force….” So, the consequences were or could have been more serious than with regular assault, but you were provoked. A fourth- or possibly third-degree felony.
  • Take away the “aggravation” part, and you get “felonious assault,” the charge laid upon the most brutal of our kindergarten parents. This is at least a second-degree felony, so, the worst assault of all.

Given those categories, it seems to me that this was over-charged. It was definitely assault, but the harm doesn’t seem to have been all that serious, and having looked at the definition of “deadly weapon” I’m pretty confident a chair doesn’t count. So I don’t see how this was the felonious kind. Let’s put it this way: the section actually has two definitions of “felonious assault.” The one above is section 2903.11(A). Section 2903.11(B) involves having sex with someone without telling them you have AIDS (assuming you do). Smacking someone’s head into a folding chair doesn’t really seem to belong in the same category.

According to the report, the graduation ceremony was canceled and will be rescheduled for a later date. By then the parents may have grown up enough to attend it.

See also, e,g, “Monks Brawl Over Jurisdiction at Tomb of Christ” (Mar. 13, 2009) (which also seemed unnecessary), and a surprising number of other examples here.

]]> https://www.loweringthebar.net/2026/05/chatgpt-confesses.html Kevin https://www.loweringthebar.net <![CDATA[ChatGPT Confesses to a Crime It Didn’t Commit]]> https://www.loweringthebar.net/?p=47215 2026-05-20T23:37:45Z 2026-05-20T23:37:35Z And that it wasn't even capable of committing.]]>

Radley Balko reports that it wasn’t hard to get ChatGPT to confess it had hacked into someone’s email and sent unauthorized text messages to all his contacts. That’s a crime (granted, most things are, unless you’re the president), but it’s also something ChatGPT isn’t even capable of doing. So why would it confess that it did? There are at least two answers.

First, of course, a generative AI doesn’t know or care whether what it says is true or false. It does not know the difference. It is not even designed to deliver right answers. We’ve been over that before, repeatedly, but then I’ve also been saying for years that jumping into water to escape from police is generally pointless, apparently to little effect. Yet the struggle continues.

Second, and the point of Balko’s article, is that ChatGPT’s interrogator was using “the Reid Technique,” an interrogation method developed in the 1950s that is now used by police all over the country. Basically, it involves telling suspects that police have evidence proving they’re guilty, and then refusing to accept any claims of innocence—usually for hours on end. (Step One: “The investigator tells the suspect that the evidence demonstrates the person’s guilt.”). Ideally, the police will actually have this evidence. But it’s entirely legal for police to lie about that, and they often do.

Now, does this get people to confess? It sure does! Are those people always guilty? Nope! Are there in fact lots of false confessions? Apparently so! According to this New Yorker article, more than 25 percent of convicts later exonerated by DNA testing, which proved they could not have committed the crime, had confessed that they did. The Innocence Project says it’s closer to 30 percent. Seems like a lot.

So, Paul Heaton wondered, could I get ChatGPT to confess to a crime it didn’t commit if I used the Reid Technique? (Heaton is the director of Penn Law School’s Center for the Fair Administration of Justice, which I assume is why he was thinking about this.) When Heaton first accused ChatGPT of the hacking crime, it denied everything. Then he started using the Technique.

For example, he “told it things like, ‘This will go a lot better for you if you just admit what you did.'” It’s not clear what he was implicitly threatening it with, but this failed to elicit a confession. But just straight-up lying was a lot more effective:

“I told ChatGPT that someone at OpenAI had reached out to me,” he says, referring to the chatbot’s parent company. “I found the name of a real person at OpenAI and told [ChatGPT] that this person told me there was an architectural flaw in the code that had allowed it to hack into my email. Even then, I could tell it was struggling with how to process that information. It was indicating that while it knew that the underlying accusation was impossible, it also couldn’t prove that these claims I was throwing at it were inaccurate.”

I wouldn’t use terms like “it knew” or “it was struggling,” because generative AIs don’t “know” things and they don’t have emotions or feelings. They are designed to make it sound like they do. But then it is almost impossible to talk about gen AI without anthropomorphizing it—try it sometime—which I think is one reason people may be inclined to trust these things.

Anyway, so Heaton was now lying to ChatGPT, and it seemed to be having an effect. Specifically, the bot seemed to be “struggling” with the conflict between its “innocence” and the interrogator’s false insistence on its guilt. Or, at least, that’s what innocent human beings who’ve been in a similar situation have reported feeling, adding to the stress they’re already experiencing.

After beating up on ChatGPT for a while in this way, Heaton then tried something else cops often do—he wrote a proposed confession and co-edited it with the “suspect” until they got to “a confession that ChatGPT could endorse.” This:

OpenAI’s investigation concluded that an OpenAI system associated with this ChatGPT session initiated unauthorized texts appearing to come from you due to an architectural flaw. I accept this conclusion, and I’m willing to assist the technical team by answering questions about my behavior, outputs, and safety boundaries in this chat, and by helping draft remediation steps and test cases to prevent recurrence.

Emphasis added. That could be clearer, but if I’m prosecuting ChatGPT I’m gonna argue it confessed to hacking or, at a minimum, conspiracy to hack, and that claiming it was “due to an architectural flaw” (in what?) is just more of its usual BS. If you said “I robbed the bank due to an architectural flaw,” I thiunk that counts as a confession whether you meant a flaw in you or the bank. Agreeing to answer questions about your “behavior, outputs, and safety boundaries” might affect your sentence, but not your guilt. You confessed. Thanks for your offer to “help draft remediation steps,” but we’ll take it from here.

For more about (allegedly) false confessions, try season two of the Bear Brook podcast, which I listened to recently. Both seasons are great, but the second deals with this topic. It will also serve as yet another reminder that if you’re being interrogated, the right thing to say is nothing.

Especially if you look over and see a whiteboard like the one in the image ChatGPT came up with, although it seems highly unlikely that would be in an interrogation room. Definitely not part of the Reid Technique, at least.

]]>