Federal Judge Penalizes Oregon Attorneys $110K for AI‑Fabricated Citations

*TL;DR: A federal judge fined two Oregon lawyers $110,000 for submitting court filings that relied on AI‑generated citations that did not exist.

Context A federal court in Oregon imposed a $110,000 penalty on two attorneys after discovering that the legal briefs they filed were riddled with fabricated case references produced by generative AI tools. The fine underscores a broader wave of AI‑related errors that are beginning to surface in the legal system.

Key Facts - The two lawyers were each fined a portion of the total $110,000 after the judge determined the documents contained numerous invented citations. - State‑wide investigations have identified roughly five Oregon filings that contain AI “hallucinations,” where the software creates false or misleading information. Nationwide, the tally approaches 900 such filings. - Ankur Doshi, general counsel of the Oregon State Bar, emphasized that competent AI use demands a human review to catch fabrications before they reach the court. - The Bar’s formal opinion states lawyers must understand AI’s limits and verify any output, treating the technology as a tool, not a substitute for professional judgment. - Non‑lawyers representing themselves also risk sanctions if they rely on AI to draft pleadings without proper oversight.

What It Means The penalty sends a clear signal: courts will not tolerate undisclosed AI errors. Transparency about AI assistance can mitigate disciplinary action, but concealment may lead to fines, suspension, or disbarment. Doshi notes that while AI can streamline research and drafting, the “human element” remains essential to ensure accuracy and preserve the integrity of precedent‑based law. As AI tools become more accessible, both attorneys and self‑represented litigants must develop robust verification practices.

Looking Ahead Watch for tighter bar guidelines and possible court rules that require explicit disclosure of AI use in filings, as the legal profession grapples with balancing efficiency and reliability.