State regulators approved more than 80 percent of requests from Berkshire Hathaway, Chubb and Travelers to exclude AI-related damages from general liability policies, according to reporting by The Information cited by PYMNTS. Florida, Connecticut and Maryland approved the highest number of those requests, and the exclusions began taking effect as early as January.

The reach of the change runs through standardized policy forms. Forms published by the Insurance Services Office underpin roughly 82 percent of United States property and casualty policies, according to an analysis by PHL Firm cited in the report. The organization introduced two optional endorsements covering bodily injury, property damage, and personal and advertising injury tied to generative AI outputs, including defamatory content and intellectual property infringement. Some carriers, including Berkley, have gone further with absolute AI exclusions across directors and officers, errors and omissions, and fiduciary liability policies.

A replacement market has formed at defined price points. Munich Re and startups including Corgi, Armilla, Mayflower Specialty and Embroker now sell standalone AI liability policies. Coverage limits run from $2 million to $50 million. Annual premiums range from a few hundred dollars to several hundred thousand dollars.

Insurers are also moving to cap AI losses inside cybersecurity policies, according to Financial Times reporting cited by PYMNTS. Policyholder Pulse noted that courts have yet to settle how broadly the new exclusions will apply.

Source: PYMNTS - https://www.pymnts.com/news/artificial-intelligence/2026/big-insurance-backs-away-from-ai-risk-and-startups-rush-in/