California Issues Historic Fine Over Lawyer’s ChatGPT Fabrications
California recently issued a historic fine to a lawyer, Mehrdad Mostafavi, for submitting court filings containing fabricated legal cases generated by ChatGPT. The state Bar of California imposed a $8,500 fine and required Mostafavi to complete legal ethics coursework after he cited six nonexistent cases in a federal court filing.
This case highlights a growing problem: the increasing use of AI in legal work and the potential for inaccuracies and “hallucinations” - instances where AI generates false information. Nicholas Sanctis, a law student at Capital University Law School in Ohio, notes that AI innovation is currently outpacing the education of attorneys, contributing to the issue.
Jenny Wondracek, who leads a project tracking these instances of AI-generated errors, anticipates the trend will worsen. She frequently encounters lawyers unaware that AI can fabricate information or who believe legal tech tools can automatically eliminate false material. “I think we’d see a reduction if (lawyers) just understood the basics of the technology,” she stated.
While instances have been documented in federal courts,Wondracek suspects state court filings contain a higher number of AI-generated fake cases,though verification is arduous due to inconsistent filing methods. She finds these errors most often among overburdened attorneys and those representing themselves, notably in family court.
The problem isn’t limited to attorneys. Wondracek has recently documented three instances of judges citing fake legal authority in their decisions.
California is now considering how to address the issue, perhaps looking to approaches taken by other states. These include temporary suspensions for attorneys who submit fabricated cases, mandatory ethics courses, and initiatives to educate law students on the ethical use of AI, such as teaching them how to avoid making similar mistakes.
Experts believe the problem will likely escalate.Mark McKenna, codirector of the UCLA Institute of Technology, Law & Policy, praised the fine against Mostafavi as a punishment for “an abdication of your responsibility as a party representing someone.” He believes the situation ”will get worse before it gets better,” due to a rapid adoption of AI by law schools and firms without sufficient consideration of appropriate usage.
UCLA School of Law professor Andrew Selbst agrees, pointing out that judicial clerks, often recent law school graduates, are being pressured to use AI to remain competitive. He notes a broader pressure felt by educators and professionals to adopt AI, stating, “This is getting shoved down all our throats…we have not yet grappled with the consequences of that.”
(This article is based on reporting from CalMatters, a Sacramento-based nonpartisan, nonprofit journalism venture.)
