Lawyers who fail to verify AI-generated content do so at their own peril – and now, potentially at the peril of fee awards. A California Court of Appeals decision adds a new wrinkle to the growing body of AI hallucination cases by asking: What happens when opposing counsel fails to detect an opponent’s fake citations?
A Familiar Story with an Unexpected Twist
In Noland v. Land of the Free, L.P., the attorney used ChatGPT to enhance his appellate briefs but failed to verify the results. The court discovered that 21 of 23 case quotations in his opening brief were fabricated, with many more fake citations in the reply brief. Some cases didn’t exist at all.
The court imposed a $10,000 sanction – standard consequences we’ve seen in similar cases. But here’s the twist: the court explicitly declined to award attorney fees to opposing counsel, despite finding the appeal frivolous.
According to the court, “respondents did not alert the court to the fabricated citations and appear to have become aware of the issue only when the court issued its order to show cause.”
Evolving Expectations for All Lawyers?
This suggests that opposing counsel might have received fee awards had they spotted and reported the fabrications. The court appears to fault counsel on two counts: failing to alert the court and failing to notice the fake citations in the first place.
The Noland decision hints that professional competence now includes heightened vigilance when reviewing any potentially AI-assisted legal work – even someone else’s.
As courts continue to grapple with AI hallucinations, one thing remains constant: human oversight and critical thinking are indispensable, no matter which side of the case you’re on.
Hat tip to Bob Ambrogi at LawSites for his coverage of this case.
Disclosure: I used Claude AI to help me develop this post, following the same critical curation and review process described in my earlier post on AI best practices.