FOLLOW US

CALIFORNIA - The Court of Appeal's Second District issued a unanimous ruling Friday that serves as both a legal precedent and stark warning to California attorneys: personally verify every citation, regardless of its source. The decision came in what should have been a routine employment law appeal, but became a cautionary tale about artificial intelligence's infiltration into legal practice.

The three-judge panel affirmed a Los Angeles County trial court's summary judgment in favor of defendants Jose Luis Nazar and Land of the Free LP in a wage-and-hour dispute with former employee Sylvia Noland. The underlying case centered on Noland's claims that Nazar failed to properly compensate her as a leasing agent and sales representative for two commercial properties.

The trial court had dismissed the suit on multiple grounds, finding that Noland was an independent contractor exempt from wage-and-hour protections and that a disputed $60,000 commission was properly withheld because the underlying transaction never closed.

But it was Noland's appeal that transformed this straightforward employment dispute into a landmark ruling on AI's role in legal practice.

The appeals court discovered that Noland's opening brief contained what it termed "extensive reliance on nonexistent legal authority." Of 23 case law quotations in the brief, 21 were completely fabricated.

"Although most of the cases to which the quotes are attributed exist, the quotes do not," the panel wrote in its published opinion. "Further, many of the cases plaintiff cites do not support the propositions for which they are cited or discuss other matters entirely, and a few of the cases do not exist at all."

The court noted that this level of fabrication "would justify striking appellant's opening brief or dismissing the appeal."

Noland's attorney, Amir Mostafavi of Mostafavi Law Group, conceded to using ChatGPT to "enhance" briefs he had written. Rather than manually proofreading the AI-generated content, Mostafavi relied on additional AI platforms including Claude, Gemini, and Grok to check for errors.

"The AI tools created fake legal authority — sometimes referred to as AI 'hallucinations' — that were undetected by plaintiff's counsel because he did not read the cases the AI tools cited," the court explained.

According to the opinion, Mostafavi told the court he was unaware that AI frequently fabricates or hallucinates legal sources. However, the panel found this lack of knowledge insufficient to excuse the misconduct.