Anthropic warns AI risks may be uninsurable — SkimNews
Get the Finance newsletter
Daily finance — markets, central banks, M&A, the prints that move money. Free.
- Anthropic disclosed in its S-1 filing that some AI-related risks may be uninsurable, as no single insurer — or the industry collectively — may have the capacity to cover catastrophic incidents involving its models
- Alissa Lugo of Gartner projects over 2,000 global 'death by AI' claims by end of 2026, defined as cases where AI played a role in a fatality, signaling rising exposure for companies
- Dennis Hendrickson of Hendrickson Insurance identifies a 'gaping hole' in cyber policies, which cover data breaches and financial loss but universally exclude bodily injury caused by AI
- Kara Higginbotham of Zurich North America confirms that most carriers are unwilling to underwrite AI-related physical harm due to unquantifiable risk, leaving a coverage gap across liability lines
- Jaymin Kim of Marsh argues AI is not a new peril but a vector for existing risks, and that isolating AI in insurance claims is impractical as it becomes embedded in everyday business processes
- Vikram Desai of Accenture draws parallels to early cyber insurance, saying AI risk is following a similar trajectory — first ignored, then excluded, and eventually priced — but at a faster pace
Why it matters: Insurers face mounting pressure to define AI liability as companies like Anthropic reveal that current policies can't cover worst-case harms. With bodily injury excluded from most cyber coverage and no standardized AI insurance, businesses across sectors risk major financial exposure from everyday AI use — a gap that could force rapid underwriting changes before regulations catch up.
Ask SkimNews
