Major AI companies, including OpenAI, Google, Meta, and Anthropic, have no insurance coverage against catastrophic AI risks despite regularly warning about the dangers of their own technologies. OpenAI reportedly carries minimal coverage of around $300 million for emerging AI risks, a figure that remains contested, while traditional insurers refuse to cover large-scale loss scenarios. Anthropic’s IPO prospectus, released in late September 2026, devotes approximately 80 of its 261 pages to potential existential risks from advanced AI, with the company’s valuation estimated at up to $2 trillion. Proposals to address this coverage gap include mandatory liability insurance for frontier AI developers, drawing from nuclear power regulations, as well as mutual insurance structures and catastrophe bonds to pool and transfer risk.
Source: Read the original article
