Yet another lawyer found out the hard way artificial intelligence is not so intelligent as a federal magistrate busted her for submitting a brief chock-full of bogus citations.
Watson & Norris represents Blake Lewis in Blake Lewis v. Entergy. Entergy filed a motion to stay proceedings. The motion generated a response in opposition to the motion on August 11. There was just one problem. The response was filled with phony citations in yet another sloppy use of AI.
U.S. Magistrate Andrew Harris cited the errors in his September 3 Show Cause order. Judge Harris said the response cited a case that was actually another case, cited a case that did not exist, and cited a Fifth Circuit holding that did not exist in the case cited. The response claimed the case, Fujita v. U.S., denied a stay when in reality it affirmed a stay. Ouch.
The Court made its feelings known:
The Court is concerned that the citations and representations detailed above have been submitted to the Court in violation of Federal Rule of Civil Procedure 11(b), Mississippi Rule of Professional Conduct 3.3, and Local Rule 83.1(c)(1). The Court is also concerned that the information submitted to the Court may be the result of insufficiently reviewed output from generative artificial intelligence (AI), which has a known propensity to fabricate case citations (so-called “hallucinations”).
Judge Harris ordered Jane Watson, the attorney who signed the brief, to submit a written response to the Court's concerns.
Ms. Watson admitted in her response she used AI to draft the memorandum. She said she had only been practicing for a year and did had never drafted a response to a motion to stay.
The attorney admitted to using a "non-firm approved" general AI service instead of using Westlaw. Her affidavit states the firm has a policy of using an AI service as long as the service is approved by the firm and all citations are checked for accuracy.
Ms. Watson filed a motion for leave to file a corrected memorandum. The Court has not scheduled a Show Cause hearing.
Unfortunately, this was not the only time Ms. Watson improperly used artificial intelligence to craft a response. Entergy filed a motion to dismiss in Dyess v. Entergy this year. The case is assigned to U.S. District Judge Carlton Reeves.
Ms. Watson filed a response to the motion on August 11. The attorney asked the Court to allow her to submit a corrected response. She admitted in her motion:
Plaintiff's counsel used AI tools to assist in drafting the memorandum response filed on August 11, 2025. Upon review, counsel identified inaccuracies in certain citations, resulting from unverified AI-generated research.
Judge Reeves ordered her to submit a corrected brief on October 1 as he allowed her to escape further embarrassment or sanctions. Ms. Watson filed the required brief.
This is not the first time Watson & Norris has gotten in trouble for the misuse of AI. U.S. District Judge Sharion Aycock is holding a Show Cause hearing for misuse of AI in Billups v. Louisville Municipal School District on October 21. Earlier post.
14 comments:
What I don't understand is why these attorneys don't double-check the work before submitting.
Hubris, arrogance, or just lazy?
I am dumbfounded that they have access to Westlaw and actually choose something that kicks out wrong info. Just how damn lazy to you have to be?
0% chance these shenanigans get past Hon. Judge Harris. He was by leaps and bounds the sharpest in our law school class.
I guess once you get used to AI doing your homework for you, it's hard to break the habit.
Filings like this are one reason my Entergy bill is so high-
Yesterday I couldn't even spell attorney; today I are one.
Young attorney. Mistakes happen. She will be a better lawyer because of this. This is a new frontier; mistakes are inevitable.
nope. not even close friend. I run a Managaed Service provider and I have over 20 attorneys of all ages and even boomers are leaning hard on the cheapest, shadiest, foreign made AI chatbots they can get. I will often see a dozen or more very sketchy browser extensions. Don’t just pin this on zoomers. It is hilarously universal. I have one client who dismissed his paralegals and wants to get rid of us just as soon as he can find an AI to manage his own IT.
The reality has been exposed. Most of the AI drivers, AI robots, and even these chatbots are actually just the same Indians who used to answer phones. They are remotely driving for Weimo, remotey walking in crowds ar events as Teslabots, and remotely “doing the needful” for our attorney friends!
Best comment on here yet!
Lynn Fitch's Policy Director has never practiced law. Are all our policies AI-generated?
@1:55. Came here to post this. I too am your classmate. I believe he was #1 from beginning to end
It's fun as an attorney who has access to, but doesn't use AI legal research to draft motions or responses. We can plug in the other big law firm party's response/motion and AI tells on itself as being untruthful/inaccurate. Silly for these firms to not double check the own software they use.
One would think that Louis Watson and Nick Watson would have better control of what is sent from their office. It is their company and their reputation.
Post a Comment