Viral Wire

AI Hallucinations in Legal Filings Lead to $110K Fines in Oregon

Over 900 fabricated court filings across US as lawyers misuse AI tools like ChatGPT.

Deep Dive

A recent $110,000 fine against two Oregon lawyers highlights a growing problem: AI hallucinations in legal filings. According to Ankur Doshi, general counsel of the Oregon State Bar, there have been approximately five identified cases in Oregon and around 900 nationally where generative AI programs like ChatGPT or Claude have fabricated cases, citations, or statements of law. The misuse extends beyond lawyers to self-represented individuals drafting pleadings with AI. Courts have responded with fines, sanctions, and threats of disbarment, especially when attorneys fail to disclose or verify AI-generated content.

The underlying tension is that AI offers substantial time savings and efficiency for legal work, but many lawyers lack understanding of how these tools function. Doshi emphasizes that the key to avoiding fabrications is human oversight—every AI-generated document needs careful review. The consequences of unchecked hallucinations go beyond individual penalties: they cost courts time and money, burden opposing counsel, and could ultimately erode the integrity of the precedent-based legal system. The Oregon State Bar's formal opinion requires lawyers to be competent in any tools they use, including understanding AI's tendency to produce false information.

Key Points
  • Two Oregon lawyers were fined $110,000 for filing court documents containing AI-fabricated cases and quotes.
  • Over 900 AI hallucination incidents have been identified in US court filings, with five in Oregon.
  • Human oversight is mandatory: lawyers must verify AI output or face sanctions, suspension, or disbarment.

Why It Matters

AI misuse in legal filings threatens the integrity of the justice system and could undermine trust in precedent-based rulings.