Generative AI hallucinations cited in 66+ court sanctions
Courts have sanctioned litigants at least 66 times for misusing generative AI tools that created false case citations, affecting both self-represented parties and licensed attorneys. The rate of sanctions has increased despite lessons from early high-profile cases, prompting courts to issue new rules and standing orders governing AI use in litigation.
What happened
U.S. federal courts have sanctioned litigants at least 66 times for misusing generative AI tools that created false or non-existent case citations. Nearly two years after the first federal court sanctions for citing hallucinated cases, the rate of such incidents has steadily increased despite the cautionary lessons.
Reported impact
Pro se litigants and licensed attorneys have faced sanctions ranging from warnings to monetary fines and dismissal of claims. At least two federal judges recently published written opinions that were withdrawn due to hallmarks of AI hallucinations, containing improper parties, factual allegations, and citation inaccuracies.
Index analysis
Courts have identified and are monitoring 232 combined local rules, standing orders, or court opinions concerning GenAI use and misuse in the courtroom. Many courts have imposed specific requirements governing GenAI usage, while others have relied on existing procedural and ethical rules to hold litigants accountable.
Governance controls
Courts have issued new standing orders, local rules, case-specific guidance, and written decisions addressing parties' reliance on GenAI tools. Legal firms are advised to improve training and regulations for using generative AI to avoid financial and reputational harm.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.