Three years ago, I wrote about one of the first instances of a lawyer citing fake cases generate by artificial intelligence (AI) in their legal briefs only to have the court call them out on it and consider sanctions. While most would think that the news that went viral in the legal community would serve as a deterrent to those tempted, it has not. Over the past three years, courts have increasingly sanctioned attorneys for submitting briefs containing fabricated citations generated by AI tools.
Most of these benchslaps involved an attorney on one side of a lawsuit. But this is a case where the court caught both sides redhanded.
In Withers v. City of Aberdeen, Tom Withers sued the city for breach of contract. The venue is the United States District Court for the Northern District of Mississippi, Aberdeen Division. Each side was represented by two attorneys with each side having one local counsel and an outside counsel admitted pro hac vice.
At one point, Aberdeen filed a motion for summary judgment, and Withers filed an opposition. When the court reviewed the filings, it could not locate some of the cited legal authorities. So the court ordered a hearing to demand explanations from both sides. Each side then asked the court’s permission to file corrected briefs. In those filings, the pro hac vice attorneys admitted to drafting the memorandums using AI and not cite checking the authorities cited. The local attorneys admitted that they did not check their co-counsel’s memorandums for accuracy before filing them.
At the hearing, all of the attorneys involved generally admitted that they failed to cite-check the memorandums they signed off on and accepted responsibility. But two of the attorneys’ explanations are worth mentioning.
One attorney was shocked at the court’s order and claimed that she did not know that AI can cite false cases and didn’t know what a hallucinated case was. The court found this explanation to be insufficient and incredulous. The court then stated that this attorney continued to file AI-generated and unverified memorandums after she received notice from the court. This resulted in another sanction from the Louisiana Bankruptcy Court for filing two memorandums with AI-hallucinated case law.
The other pro hac vice attorney involved stated that she used AI software that was built for in-house legal research, only used it for 90 days, and that it was custom made to produce results from the state where she practices but not from Mississippi. It is not clear whether the term “in-house legal research” means an office-use legal research tool or if it is software specifically tailored to in-house corporate attorneys. She also admitted to this being her first time handling a Mississippi case.
After considering the facts and the explanations from the attorneys, the court imposed sanctions on all of the attorneys involved. For the two local counsel who simply signed off the memorandums, they were each ordered to pay $1,000. The two pro hac vice counsel were sanctioned $2,500 and $3,500.
Also, every attorney involved has been disqualified from the case. The pro hac vice attorneys have had their admissions revoked and are banned from seeking admissions to the court for two years. The court will also send a copy of its sanctions order to the state bars where each of the attorneys are practicing where they will decide the appropriate response.
The judge also cancelled the trial.
While this could have been an interesting way to pit AI vs. AI to see which is better at motion drafting, it remains that AI cannot cite cases correctly even after several years. If the general public sees this, they will not only lose more respect towards the profession, but they will try their luck with AI and use it to conjure up more lawyer jokes.
I did not name the attorneys involved for several reasons. First, being mentioned on Above the Law leaves a big mark on the internet, and I think the professional repercussions they suffered now and possibly later is enough. The important lesson is that courts and state bars will get tougher on attorneys who do not wisely use AI. Second, I would like to know why some attorneys are taking a risk on their professional reputation by submitting AI-generated legal memoranda without verifying accuracy. I would like to hear and anonymously share their stories so that others will learn from their mistakes. I think it is not enough to attribute their behavior to pure laziness. If you want to share your (or your “friend’s”) story, please send me an email.
Gavel Bang: Rob Freund for finding this gem.
Steven Chung is a tax attorney in Los Angeles, California. He helps people with basic tax planning and resolve tax disputes. He is also sympathetic to people with large student loans. He can be reached via email at stevenchungatl@gmail.com. Or you can connect with him on Twitter (@stevenchung) and connect with him on LinkedIn.
The post Court Sanctions Lawyers From Both Sides In The Same Lawsuit For Filing Briefs With AI-Hallucinated Cases appeared first on Above the Law.
Three years ago, I wrote about one of the first instances of a lawyer citing fake cases generate by artificial intelligence (AI) in their legal briefs only to have the court call them out on it and consider sanctions. While most would think that the news that went viral in the legal community would serve as a deterrent to those tempted, it has not. Over the past three years, courts have increasingly sanctioned attorneys for submitting briefs containing fabricated citations generated by AI tools.
Most of these benchslaps involved an attorney on one side of a lawsuit. But this is a case where the court caught both sides redhanded.
In Withers v. City of Aberdeen, Tom Withers sued the city for breach of contract. The venue is the United States District Court for the Northern District of Mississippi, Aberdeen Division. Each side was represented by two attorneys with each side having one local counsel and an outside counsel admitted pro hac vice.
At one point, Aberdeen filed a motion for summary judgment, and Withers filed an opposition. When the court reviewed the filings, it could not locate some of the cited legal authorities. So the court ordered a hearing to demand explanations from both sides. Each side then asked the court’s permission to file corrected briefs. In those filings, the pro hac vice attorneys admitted to drafting the memorandums using AI and not cite checking the authorities cited. The local attorneys admitted that they did not check their co-counsel’s memorandums for accuracy before filing them.
At the hearing, all of the attorneys involved generally admitted that they failed to cite-check the memorandums they signed off on and accepted responsibility. But two of the attorneys’ explanations are worth mentioning.
One attorney was shocked at the court’s order and claimed that she did not know that AI can cite false cases and didn’t know what a hallucinated case was. The court found this explanation to be insufficient and incredulous. The court then stated that this attorney continued to file AI-generated and unverified memorandums after she received notice from the court. This resulted in another sanction from the Louisiana Bankruptcy Court for filing two memorandums with AI-hallucinated case law.
The other pro hac vice attorney involved stated that she used AI software that was built for in-house legal research, only used it for 90 days, and that it was custom made to produce results from the state where she practices but not from Mississippi. It is not clear whether the term “in-house legal research” means an office-use legal research tool or if it is software specifically tailored to in-house corporate attorneys. She also admitted to this being her first time handling a Mississippi case.
After considering the facts and the explanations from the attorneys, the court imposed sanctions on all of the attorneys involved. For the two local counsel who simply signed off the memorandums, they were each ordered to pay $1,000. The two pro hac vice counsel were sanctioned $2,500 and $3,500.
Also, every attorney involved has been disqualified from the case. The pro hac vice attorneys have had their admissions revoked and are banned from seeking admissions to the court for two years. The court will also send a copy of its sanctions order to the state bars where each of the attorneys are practicing where they will decide the appropriate response.
The judge also cancelled the trial.
While this could have been an interesting way to pit AI vs. AI to see which is better at motion drafting, it remains that AI cannot cite cases correctly even after several years. If the general public sees this, they will not only lose more respect towards the profession, but they will try their luck with AI and use it to conjure up more lawyer jokes.
I did not name the attorneys involved for several reasons. First, being mentioned on Above the Law leaves a big mark on the internet, and I think the professional repercussions they suffered now and possibly later is enough. The important lesson is that courts and state bars will get tougher on attorneys who do not wisely use AI. Second, I would like to know why some attorneys are taking a risk on their professional reputation by submitting AI-generated legal memoranda without verifying accuracy. I would like to hear and anonymously share their stories so that others will learn from their mistakes. I think it is not enough to attribute their behavior to pure laziness. If you want to share your (or your “friend’s”) story, please send me an email.
Gavel Bang: Rob Freund for finding this gem.
Steven Chung is a tax attorney in Los Angeles, California. He helps people with basic tax planning and resolve tax disputes. He is also sympathetic to people with large student loans. He can be reached via email at stevenchungatl@gmail.com. Or you can connect with him on Twitter (@stevenchung) and connect with him on LinkedIn.
The post Court Sanctions Lawyers From Both Sides In The Same Lawsuit For Filing Briefs With AI-Hallucinated Cases appeared first on Above the Law.
