US | Incident Report
Hallucination
2026-01-26
Federal judge sanctions two lawyers for AI-hallucinated case citations in filing
AI Model: Unspecified generative AI tool HallucinationsCourt SanctionsRule 11Professional Responsibility
I. Executive Summary
A U.S. federal judge in the Eastern District of Pennsylvania sanctioned two attorneys after identifying multiple AI-related “hallucinations‘ in briefing they signed. One attorney received a monetary sanction, while the other received non-monetary directives tied to sharing updated AI-use policies. The court declined disciplinary referral, emphasizing deterrence and the non-delegable duty to verify authorities in signed filings.
II. Key Facts
- Court identified multiple fabricated/misapplied authorities and inaccurate quotations described as AI hallucinations.• Monetary sanction of $4,000 imposed on one attorney; additional non-monetary measures ordered for both.• Court required dissemination of updated AI-use policies to professional groups as a deterrent measure.• Judge declined bar/disciplinary referral conditioned on compliance with the order.
III. Regulatory & Ethical Implications
Reinforces that courts will treat AI-assisted citation errors as Rule 11 / candor failures and may impose tailored sanctions emphasizing verification, supervision, and local counsel oversight; increases malpractice and discipline exposure for AI-assisted drafting failures.