The scope of coverage gaps remains unsettled. Insurers have not yet standardized how they will allocate liability when AI systems cause harm, particularly where responsibility is distributed across vendors, model providers, and end-user businesses. Courts and regulators have not established clear precedent for AI-caused losses, leaving underwriters to price risk in an environment of genuine uncertainty about future legal exposure.
Companies deploying generative AI and embedded tools across hiring, lending, customer service, and security operations now face a timing problem: they are operationalizing AI while insurers are simultaneously narrowing or redefining coverage. Risk managers should audit their current policies for AI exclusions, document their AI governance and monitoring practices, and engage underwriters early about coverage scope. This is not a future issue. The protection gap is opening now.