Berkshire, Chubb, and Travelers Are Removing AI Coverage

Key Takeaways

• Berkshire, Chubb, Travelers win AI exclusion approvals in 80% of filings.
• AI liability exclusions target discrimination, IP, and autonomous system damages.
• 140% YoY rise in AI lawsuits drives carriers to carve out coverage.
• Deloitte forecasts $4.7B annual AI liability premium market by 2032.
• Brokers must identify AI gaps and recommend specialty products now.

Pulse Analysis

The wave of AI exclusion endorsements marks a watershed moment for commercial lines. Regulators in Florida, Connecticut, and Maryland have cleared the majority of filings, effectively stripping AI‑related perils from traditional general‑liability, errors‑and‑omissions, and directors‑and‑officers forms. Insurers cite a dramatic uptick in AI‑generated claims—court filings rose 140% in 2025 and a MIT‑backed study recorded a 978% increase in generative‑AI lawsuits since 2020. By isolating these exposures, carriers echo the early cyber‑insurance playbook, where blanket coverage gave way to bespoke products as risk became better understood.

Analysts estimate the nascent AI liability market could generate $4.7 billion in annual premiums by 2032, according to Deloitte. First‑mover carriers that launch dedicated AI policies can command premium‑price discovery, leverage surplus‑line capacity, and build underwriting expertise before pricing stabilizes. The competitive landscape is already shifting: Armilla AI, Testudo Global, and Munich Re have introduced niche solutions, while the three large carriers are poised to expand their specialty footprints. The strategic calculus for insurers now hinges on balancing capital allocation for new product development against the risk of underwriting gaps that could erode loss ratios.

For brokers and product managers, the imperative is immediate. Every commercial account deploying AI agents, autonomous equipment, or AI‑generated content must undergo a gap analysis at renewal, as the exclusions are already in force. Firms should map state‑by‑state approval data, align it with client exposure profiles, and partner with surplus‑line carriers offering AI‑specific endorsements. Legal teams must also audit existing policy language to ensure no implicit AI coverage remains. As AI models like Claude Mythos demonstrate the ability to autonomously discover software vulnerabilities, the intersection of cyber and AI liability will intensify, making integrated risk solutions a critical differentiator for forward‑looking insurers.