Insurers Warn: AI May Soon Be Too Risky to Cover

What happens when the technology everyone is rushing to adopt becomes too risky for insurers to cover? According to the Financial Times, we may be about to find out.

Several major insurers, including Great American, Chubb, and W. R. Berkley, are reportedly seeking approval from U.S. regulators to exclude broad AI-related liabilities from corporate insurance policies. One underwriter told the FT that the outputs of AI models are “too much of a black box” to insure confidently.

AIG, also mentioned in the report, clarified to TechCrunch: “AIG was not specifically seeking to use these exclusions and has no plans to implement them at this time.”

The caution is understandable. AI errors are already creating costly, real-world problems. In March, Google’s AI Overview wrongly claimed a solar company faced legal troubles, sparking a $110 million lawsuit. Air Canada had to honor a discount invented by its chatbot last year. And fraudsters used a digitally cloned executive to steal $25 million from London engineering firm Arup in what appeared to be a legitimate video call.

For insurers, the real fear isn’t a single large payout—it’s the systemic risk of a flood of simultaneous claims if a widely adopted AI model fails. As one Aon executive explained, covering a $400 million loss to a single company is manageable. But handling thousands of losses triggered by a single AI malfunction? That could overwhelm the industry.