Anthropic liability warnings sharpen UK insurers’ coverage challenge

Anthropic has warned in its IPO prospectus that its technology could pose “catastrophic or existential risks to humanity” without appropriate safeguards. The Claude developer also acknowledged uncertainty about where and how it could be held liable for its systems’ actions, bringing questions about insurance coverage into a formal legal disclosure.

The company identified potential regulatory action over AI-related harms, including suicide and child safety, as a material risk. It also highlighted cybersecurity incidents involving increasingly autonomous AI agents. Separately, owners of Elvis Presley’s music sued Anthropic over the allegedly unauthorised use of his songs to train Claude, underscoring that training-data disputes are already reaching courts.

The disclosures coincided with a joint Bank of England and FCA survey finding that 86% of UK financial services firms use AI, with the median insurer running 18 AI systems in production. As adoption expands and the FCA positions itself as an “agentic regulator,” insurers face pressure to clarify policy language, accountability and pricing for risks whose legal boundaries remain uncertain.