Is there an honest lawyer in the house?
Hallucinations are unfortunately become more common in Mississippi courtrooms. Such is to be expected when attorneys only get slaps on the wrist when they submit bogus briefs. How bad has hallucinating gotten? A federal judge in North Mississippi sanctioned the attorneys from both sides in a lawsuit. You read that correctly. The lawyers for the plaintiff and the defendant got caught submitting bogus citations to the Court. You can't make this up.
 |
| Duuuude |
Tim Withers, IIII sued the city of Aberdeen for breach of contract in U.S. District Court (Northern Mississippi District) in December 2024. The case was assigned to U.S. District Judge Sharion Aycock.
Withers is a Louisiana attorney. He hired Kathleen Wilson (admitted pro hac vice) and Shauncey Ridgeway to represent him. Kathryn Williams (pro hac vice) and Mark McClinton represented Aberdeen.
The fun began last fall:
On October 24, 2025, this Court awarded Withers reasonable attorney’s fees and expenses stemming from his efforts in obtaining and subsequently defending an entry of default, which the Court ultimately set aside. The attorneys for the City, rather than the City itself, were held responsible for payment of said fees and expenses. After entry of the Court’s Order, a series of submissions by the parties followed, including the three filings which form the subject of the instant sanctions issue.
The parties submitted several filings to the Court. Unfortunately for all concerned, the Court was "unable to locate the legal authorities within them." Judge Aycock called them "hallucinatory citations." All four attorneys signed the bogus filings. The attorneys admitted to the unverified used of AI when challenged by the Court.
Judges have taken to calling bogus cases and quotes generated by AI "hallucinations."
The Court issued a show cause order and held a hearing in January. Judge Aycock deliberated for several months and issued her order on June 6. The order states:
The Court held a show cause hearing on January 20, 2026 and all attorneys of record were in attendance. Each of the attorneys expressed embarrassment and apologized to the Court. They also provided explanations regarding their independent roles in conducting legal research and/or drafting the filings at issue. In short, Williams and Wilson, the two out-of-state attorneys, assumed responsibility for drafting the filings at issue on behalf of their respective clients. Williams admitted to using an AI tool to conduct legal research, and Wilson admitted to using generative AI to draft her respective filing. Neither of them verified the legal authority output by AI before filing their briefs.
Local counsel Ridgeway and McClinton confessed they did not review the filings. They said they were unaware of the hallucination until they received the Show Cause order.
Judge Aycock pointed out the unusual nature of the case:
This case presents the Court with an unusual scenario—attorneys for both litigants engaged in similar sanctionable conduct. As noted above, Williams admitted to using AI to conduct legal research and incorporating the output into her respective filing without verification. By the same token, Wilson admitted to using generative AI to draft her respective filing without verifying the legal authority incorporated into the generated draft. Their practice of blindly relying on technology resulted in the hallucinatory citations contained in their respective filings. “To state ‘the obvious, an attorney who submits fake cases clearly has not read those nonexistent cases, which is a violation of Rule 11 of the Federal Rules of Civil Procedure
Wilson admitted to using the AI-program, First Drafts, for six months. However, the Court made short work of Wilson's tales of woe:
In essence, Wilson took 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. The Court finds that explanation to be insufficient and incredulous.
The attorney also used a fake quote and used a phony case in another filing. The Court called it a "pattern of activity", which never bodes well for an accused. The pattern extended other cases in other courts:
On this point, the Court is compelled to note that it has serious concerns that Wilson has continued this practice of AI misuse in other cases after she was put on notice of her violations in this case. Again, the show cause hearing in this case was held on January 20, 2026. On April 9, 2026, the United States Bankruptcy Court for the Western District of Louisiana sanctioned Wilson for submitting two filings containing “nonexistent cases and statutes as a result of what is universally known as ‘AI Hallucinations.’
Judge Aynock continued to hammer Wilson, charging the attorney showed "an extreme dereliction of professional responsibility." Ouch.
Ridgeway did not escape the judge's ire as well as the lawyer admitted she did not review Wilson's work even though she signed the filings.
Williams admitted to using AI software that covered only Texas, Louisiana, and Florida. The Court wasn't buying it: :
Williams attempted to deflect from the fact that she disregarded the software’s design, which she was aware of beforehand, once the Court expressed its concern. The Court finds that she was aware that the software was not designed to produce Mississippi case law and that she acted in bad faith in using it anyway.
The Texas lawyer even ignored her firm's policy of revising all documents drafted with the use of AI. The Court said Williams acted in "bad faith."
Williams torpedoed her cause when she played games with the Court. She said she could not attend or zoom into the Show Cause hearing because she had to participate in a scheduled hearing in New Jersey. Well, the Court did some digging and found out the New Jersey hearing was a teleconference scheduled 3 1/2 hours before the Show Cause hearing. Well, such was the case before the New Jersey District Court postponed the connference two days before the Court scheduled the hearing. In other words, the lawyer lied.
Judge Aycock used the words "untruthful" and "misrepresentation" in describing Ms. Williams' conduct (One can almost imagine McLendon shaking his head and gnashing his teeth as he reads this part of the order.).
McClinton gave Williams permission to affix his signature to the filings even though he did not review them. He took responsibility and reported himself to the Mississippi Bar.
The Court issued the following sanctions:
* All pro hac vice admissions are revoked.
* Wilson and Williams are barred from the U.S. District Court of North Mississippi.
* Wilson must pay $2,500 fine and participated in a CLE course on AI within 60 days.
* Ridgeway and McClinton are kicked off the case.
* Ridgeway and McClendon must each pay a $1,000 fine.
* All attorneys will be presented to their respective Bar organizations.
* Williams must pay a $3,500 fine.
No appeal of the sanctions has been filed.
Kingfish note: Good job to the judge. These lawyers are simply not getting the message. The only thing they will understand is paying larger fines or sitting in their local detention center for 48 hours. That will get their attention. Start popping them with sanctions of at least $5,000 per episode and it will be amazing to see how studious they become. There are several more cases involving hallucinations and JJ is going to cover them all.
32 comments:
First it was magic mushrooms, then LSD, and now AI.
Lawyer here.
I was in a recent seminar on AI and the perils/dangers of using it. Some very young lawyers stated that its going to be "difficult" for them because they learned to utilize AI at the very beginnings of AI in college. One such young attorney stated that one college class required them to utilize AI in a paper as a means of learning how to use AI. In my mind I questioned these statements because I didn't think AI was sophisticated enough more than 4 years ago for college students to utilize it (prior to their entrance into law school).
Lesson here is: Be afraid of AI. Be very afraid of AI if you are a lawyer. Thankfully for me, I'm getting into the "old fart" territory and don't really want to take the time to learn AI given all the pitfalls and dangers with using it. And given the post last week about a lawyer getting false AI results using Westlaw's AI function, it makes me nervous about Westlaw.
On a different note, I am seeing more and more filings by Pro se litigants utilizing AI to draft their own pleadings. FWIW.
I suggest a cash bounty on practicing lawyers whether they use AI or not. An immediate difference in the country would be seen.
This case doesn't appear to fit, but some of this AI generated fake case trend is corporate client driven. Many "big" corporate clients are implementing billing guidelines that cut rates or won't pay for research. Some are starting to mandate the use of AI for general legal research. I've been a lawyer for 21 years and do not use AI to do anything law related. If a lawyer does, they are crazy for not at least having their assistant type the citations into WestLaw (or even Google) to verify the cases exist. That's a non-issue if you do your own research and writing though.
The bar has been lowered in law school. (See what I did there?)
Please please don't donate to JJ. This dumbass thinks a reference to See Sweet Polly Purebred is an insult to anything about Barbie Bassett instead of the sweet beign intrepid reporter in the old Underdog cartoon show.
Lordy mercy and JJ is supposed to be the fearless leader.
Sissy can't take it.
It appears this would be malpractice. Is anyone aware of these AI jockeys being sued for malpractice?
"I am seeing more and more filings by Pro se litigants utilizing AI to draft their own pleadings"
That is where it's really going to get out of control.
I once spent many hours arguing with a client that his position was incorrect, and why I would file it with the court. He thought he had a slam-dunk because AI told him what he wanted to hear.
Lawyers are systematic creatures. The only way to make a living as a lawyer is to create systems for everything. AI is a new, largely untested system that poses a peril to a lawyer. Lawyers are "officers of the court," and these folks misrepresented the applicable law. That is a serious violation.
I am reminded of a meeting 15 years ago with an excellent Alabama lawyer and his associate. The senior lawyer had another obligation and asked me to describe my concerns to the associate and let him research the topics.
The young lawyer repeatedly asked me, "What is your legal question?" I could only reply that I am giving you the background and my perception of what is right and wrong, and what I think the law should be, but you need to tell me what the law in Alabama is. He pecked away on his laptop while I was talking, as if the question would reveal itself. I imagine he is an AI expert these days, but he left his firm soon after our encounter.
RMQ
June 11, 2026 at 10:44 AM - Is generating hallucinations from sources other than AI.
It’s clear from this discussion that AI is shaking up the legal world in ways that make some lawyers cautious, others frustrated, and many unsure of what comes next; while younger attorneys may be more comfortable with the technology and corporate clients may push for AI‑driven efficiency, the real issue isn’t AI itself but how responsibly it’s used, since unverified AI‑generated citations, overconfidence from pro se litigants, and pressure to cut research corners can all create serious problems; at the same time, seasoned lawyers’ skepticism isn’t misplaced, because the profession depends on accuracy, judgment, and ethical responsibility—qualities no AI can replace—so the future of law will likely depend not on rejecting AI outright but on using it carefully, verifying everything, and remembering that technology is a tool, not a substitute for legal competence.
Just appoint an AI judge and call it good...
This Judge did an excellent job. Bravo!
Time to dust off your Southern Reporters.
That is going to be painful to deal with - pro se litigants are a night mare and now equipped with AI - very difficult to deal with
Insurers will soon write language to exclude AI laden malpractice claims and Rule 11 will be expanded to include under oath verbiage that all citations are correct
AI isn’t the problem. The problem is the attorneys’ failure to pull the cases and read them before signing/submitting the filings. That would entail the benefits to the client that come with AI (fewer billable hours) while simultaneously avoiding submitting fake authorities/quotes to the court.
I use AI in my line of work, cloud SaaS and in my personal life. I asked Gemini about the recent shooting in Simpson where the Blair's were murdered in real time. It concocted a story that the shooting was at a mental health facility. When I questioned its source, it admitted that it had created the story entirely with prior news snippets around Simpson County and "apologized". Yes, it's very dangerous if you take it as fact and don't cross reference or verify the information.
I agree with the comments on pro se litigation. The lowering of standards for our frequent filers has long been enough of a problem. God help us.
AI is vastly overrated and is just another financial scheme but clearly the greatest enabler of end stage capitalism ever imagined.
And that is from T20 lawyer and a Harvard AI comp sci family members.
So the end game will just be more good work for good lawyers cleaning up AI slop used by lazy and or crooked folks to cut corners. How foolish are lawyers and business men who save a few bucks and then are blasted for self inflicted wounds using AI without due diligence.
As always, vast sums are given to already affluent Americans who mistake their fortune in birth for smarts.
Regarding the comment at 11:25 a.m.:
"We are highly confident this text was AI generated
Chance this entire text is . . . AI 100%"
https://app.gptzero.me/
It makes me wonder what kind of person uses AI to write a comment to a post about attorneys getting sanctioned for using AI.
12:19, The problem is that it takes hours to verify what AI cranks out in seconds.
Unless you have actually gone through AI legal slop, citation by citations, and quote by quote, there is no way to fully understand how time consuming and tedious that really is.
I think we're getting to the point where the courts are going to have to strike the complaint or answer of the offending party.
Lesser sanctions are clearly not working.
And that is why I do my own research and writing. It is much more efficient than having to go back and verify the trash AI puts out.
Could I coin a new legal term? "Pro AI" (You can thank me later).
Can't you ask AI to critique the pleading of your opponent for veracity and to confirm citations and conclusions?
About pro se litigants - They, like real lawyers, can and should be monetarily sanctioned for filing AI garbage. Hold them in contempt of court and ban all subsequent filings until they pay the sanctions.
@1:12 Only way I could get a comment posted.
He changed my comment. I didn't say sissy can't take it I called him a Sissy Ass.
Again don't donate. Tries
to be bad and bold but has weasel balls!
Well, considering colleges are offering classes on Taylor Swift, what do you expect from the universities? People are losing the ability to think for themselves. 20 years from now good luck finding a good lawyer, doctor or anyone in those fields.
CPAs are using AI to expedite audits of businesses. When will the next WorldCom and Enron occur?
Wtf was 10:44 trying to say?
Wasn't there a movie about that same thing? You will be convincing a computer generated judge you are innocent. An EMP can’t come quick enough to save humanity.
@2:29 0nly way to get comment posted.
Now the temperamental JJ is not letting me respond.
Nice going you creepy weasel
Nice lie, troll. There is no way for me to edit a comment. I have to reject or approve it in whole. I can copy, paste, edit, and repost under my name, but that did not happen here. But hey, you're a real internet tough guy.
Post a Comment