AI risks make some insurers wary of corporate liability

Insurance firms are hesitant to cover potential financial damages linked to artificial intelligence, according to a recent report from RAND Corporation. The think tank identified a range of AI-related risks, such as incorrect outputs, deepfakes, privacy breaches, intellectual property disputes, fraud, product defects, and biased decisions.

These emerging challenges require a new approach to insurance, as existing coverage options are inadequate. Companies are concerned about the financial implications of deploying unreliable AI systems, but insurance providers are cautious about assuming this risk. If AI becomes uninsurable, the entire industry may have to scale back its ambitions and sales targets, while corporate customers may delay AI projects to adhere to their fiduciary duties.

RAND's report highlights the discrepancy between the rapid adoption of AI by businesses and the fragmented market for insuring or not insuring AI usage, particularly in the United States. Insurance firms are proactively excluding AI-related liability from their products, such as W. R. Berkley's D&O, E&O, and Fiduciary Liability insurance offerings.

During their Q4 2025 earnings call, Berkley emphasized the importance of underwriters comprehending the impact of AI on their insureds and understanding the associated risks to control and price them appropriately. Some insurance carriers have started incorporating optional language into their policies to exclude AI-related harms, such as bodily injury, property damage, and other damages caused by generative AI.

However, RAND insists that policy researchers, brokers, carriers, and reinsurers must establish a common taxonomy to track AI incidents and claims, and state regulators should develop an AI Coverage Notice to clarify coverage and exclusions.