The ongoing challenge of AI-induced errors in legal documentation has taken a novel turn with a recent decision from the California Court of Appeals. In the case of Noland v. Land of the Free, L.P., the court not only called attention to the issue of fabricated AI-generated citations but also broached the responsibility of opposing counsel in detecting such inaccuracies.
This decision comes in response to an attorney, Amir Mostafavi, who used AI tools like ChatGPT to augment his appellate briefs, leading to numerous inaccurate and non-existent citations. Despite imposing a $10,000 sanction on Mostafavi, the court notably refrained from awarding attorneys’ fees to the opposing party, the reasoning being their failure to catch and report these fabrications. The court suggested that the timely alerting of falsified citations could potentially merit rewards for opposing counsel, highlighting an emerging expectation for heightened vigilance in legal practices.
The particulars of the case fit within a pattern of increasing scrutiny over AI’s role in legal proceedings. Mostafavi’s utilization of AI—producing 21 out of 23 fabricated citations in one brief alone—drove the court to sanction not only for the failure to cross-check AI outputs but also to enforce a professional duty towards accuracy and integrity in legal documentation.
This ruling is significant as it is potentially the first to probe the duty of lawyers to detect and report their opponents’ erroneous AI outputs—a duty traditionally limited to self-scrutiny. The denial of attorneys’ fees to the opposing team further reflects a shift towards expecting lawyers to act as gatekeepers against AI inaccuracies.
The court’s approach might imply a broader principle that detecting and reporting AI errors could become a component of legal professional competence. As cases of AI-induced errors in legal documentation proliferate, the imperative for lawyers to rigorously check the validity of citations may grow. As articulated in a separate analysis of the broader landscape of AI citations in legal contexts, similar incidents have fuelled discussions at various judicial levels, suggesting an industry-wide pivot towards codifying such vigilance as part of a lawyer’s standard duty.
With courts yet to establish clear guidelines, the legal community might soon find itself adapting to these evolving expectations, as the ease of generating fake legal citations via AI necessitates new levels of diligence and professional responsibility.
For more details on the case and evolving legal standards around AI usage, visit the full article here.