AI Risks Invisible to Insurers

AI presents a significant blind spot for the insurance industry, as many insurers are exposed to risks from AI systems embedded within traditional policies. A report indicates that over 90% of an insurer's exposure to AI agents resides in "silent" cover, meaning these risks are largely unpriced and often unknown to the insurers themselves. This situation is exacerbated by a shift from AI producing content to AI agents taking autonomous actions.

These evolving AI risks can lead to a wide array of claims, including professional negligence, cyber attacks, and even wrongful death. Recent incidents highlight these concerns: Google faces an $110 million defamation lawsuit due to its AI Overviews, and Air Canada must honor a chatbot-generated discount. A UK engineering firm, Arup, lost $25 million due to an AI-powered deepfake fraud.

The potential financial implications are staggering. Experts estimate that a major AI disaster could result in approximately $100 billion in direct damage, with broader economic impacts potentially reaching trillions of dollars. This could occur if insurers withdraw coverage, leading to slowed AI adoption and investor retreat, drawing parallels to the post-9/11 terrorism insurance market collapse.

To address this, the report advises insurers to develop dedicated AI coverage rather than simply excluding AI risks from existing policies. This should be supported by new underwriting standards and technical auditing that can keep pace with rapidly evolving AI models. The severity of AI-related incidents is escalating, moving from mere policy hallucinations to significant real-world harm, including wrongful death and unauthorized legal practice.