2026-09-17 · Source: Insurance Business UK
Summary in 3 Points • Beazley's AI Clarifying Endorsement covers AI-driven cyber attacks but not clients' own AI use • The Insurance Services Office introduced generative-AI exclusions for liability policies effective January 1 • Financial Times reported Beazley drafting policy language to cap AI-related payouts for regulatory breaches --- Beazley has introduced an AI Clarifying Endorsement for its cyber policies, explicitly covering AI-driven cyber attacks, such as those utilising large language models for reconnaissance and phishing. This endorsement addresses the gap concerning attacker-side AI use but does not clarify coverage for a client's own AI use, such as chatbots or internal models. Historically, cyber policies have covered AI-related losses by omission, leading to inconsistencies in claims handling. The Insurance Services Office (ISO) has responded to this ambiguity by introducing generative-AI exclusion endorsements for commercial general liability policies, effective from January 1. For the **London Insurance Market**, Beazley's move to affirm coverage for AI-driven attacks is significant, as it aligns with efforts to eliminate "silent AI" risks, similar to the earlier "silent cyber" issue. This development is particularly relevant for **cyber insurance** lines, where London syndicates, including Beazley, lead a substantial share of the global market. The Financial Times' report on Beazley drafting policy language to cap AI-related payouts for regulatory breaches, with sub-limits at 10% of policy value, highlights the ongoing debate over AI coverage limits, which could influence underwriting strategies in the London Market. **Underwriters** and **brokers** should consider the implications of Beazley's endorsement and ISO's exclusions when drafting and renewing policies. The explicit naming of AI-driven attack coverage could streamline claims processes and reduce disputes. However, the lack of clarity on a client's own AI use necessitates careful policy wording to avoid potential coverage gaps. **Risk managers** should ensure that their clients' AI applications are adequately addressed in their insurance programmes to mitigate exposure to AI-related liabilities.