Four Mississippi lawyers sanctioned after AI hallucinations expose verification failures

密西西比四律师同遭制裁：AI 幻觉案扯出律师行业的核验失职

In June 2026, a Mississippi federal court ejected and fined all four lawyers on both sides of a case after each assumed someone else had verified AI-generated fake citations. The out-of-state plaintiff's attorney used AI to fabricate precedents; the out-of-state defense attorney knowingly applied Texas-specific legal software to a Mississippi case; both local lawyers signed filings without reading them. The judge revoked pro hac vice admissions and imposed a two-year ban from the district. This is not an isolated incident—239 U.S. lawyers have been sanctioned for AI hallucinations as of February 2026, with 712 global rulings in 2025 alone. Stanford testing shows legal-specific tools like Lexis+ AI and Westlaw AI-Assisted Research still hallucinate at rates of 17% to 34%, and vendor claims of 'hallucination-free' remain unsubstantiated. Courts are clear: Rule 11 has always required reasonable inquiry upon signing. AI only swaps the delegate from human to tool—the verification duty is unchanged. 79% of U.S. lawyers now use AI, but 43% of firms have no usage policy and 54% offer no training. When execution cost drops to near zero, judgment and verification become more expensive.

Why it matters: Hits all three HKR axes with hard data, not vague warnings. Capped at 78 because it's a vertical industry case, not an AI tech development.