Insurers Move to Exclude AI Claims, Leaving Fintechs and Banks Scrambling for Coverage
Major commercial insurers are quietly rewriting the rules of corporate liability coverage, seeking permission from regulators to carve out losses tied to artificial intelligence. The filings, which involve household carriers, would allow insurers to deny claims that stem from the use, development, or deployment of AI tools, including chatbots and autonomous agents. This shift has already forced banks and fintech firms into an urgent reassessment of their risk programs.
Why insurers are changing course
Underwriters say the math no longer works for traditional policies when faced with the prospect of widespread, simultaneous losses triggered by a single algorithm or vendor. Regulators and industry data providers have rolled out optional endorsement forms that let carriers explicitly exclude generative AI exposures, with some language effective on January 1, 2026. The new forms give insurers a clear legal basis to limit or refuse payment for harms that arise from model outputs.
Trigger events that hardened positions
A series of high profile incidents over the past two years crystallized the problem for insurers, showing how quickly an automated output can produce outsized liability. Examples include product descriptions or summaries that falsely implicated companies in litigation and social engineering attacks that used digitally cloned voices. Those episodes helped convince underwriters that an isolated failure could cascade into very large, correlated claims.
What it means for banks and fintechs
The practical outcome for regulated financial firms is immediate and painful. Renewal negotiations are becoming contested, with carriers offering narrower coverage, higher premiums, or sub-limited protection for AI-related exposures. Some market wording already applies a de facto cap, so a cyber policy with a $5 million overall limit might only provide about $250,000 for an LLM-related loss, exposing firms to the remainder. Smaller lenders and payments companies that rely on third-party AI services are particularly vulnerable.
Market responses and workarounds
Brokers and specialist underwriters are racing to fill the gap. A handful of new insurers and Lloyd's coverholders are launching tailored policies that explicitly insure certain classes of generative AI risk, such as professional indemnity for model outputs, intellectual property claims arising from generated content, and regulatory fines. At the same time, traditional brokers are advising clients to build detailed inventories of where AI sits in business flows, and to strengthen vendor contracts and oversight as a condition for negotiating coverage.
Near term outlook
Insurers emphasize that many filings are precautionary and may not be implemented immediately. Nevertheless, the regulatory approvals being sought create optionality that carriers can flip on as claim activity grows. The net effect is a clearer signal to banks, fintechs, and their boards: deploying AI without a documented risk control and insurance strategy risks leaving losses on corporate balance sheets. The insurance market is moving from broad cushion to narrowly defined walls, and firms that do not adapt will likely face both higher outlays and tougher conversations with regulators and customers.