AI insurance gap threatens all companies, Anthropic warns

Anthropic warned in its IPO filing that insurance may be insufficient to cover its AI-related liabilities, potentially exceeding the industry's capacity. The company cited risks from IP theft to existential threats, leaving substantial uninsured losses possible. The filing notes AI models may act unintentionally, causing liability that current policies cannot adequately mitigate. Experts say bodily injury claims from AI incidents remain a major coverage gap, with cyber policies typically excluding such harm. Gartner projects "death by AI" claims could exceed 2,000 by end of 2026. Industry observers compare AI liability to early cyber insurance, predicting insurers will eventually set standards through exclusions and separate policies. Most current claims involve IP disputes, while broader coverage remains nascent and uneven across business lines.