AI is too risky to insure, say people whose job is insuring risk
What happens when the software everyone is eager to adopt becomes too risky for anyone to insure? According to reporting from the Financial Timeswe’re going to find out.
Major insurers, including AIG, Great American and WR Berkley, are asking US regulators for permission to exclude AI-related liabilities from corporate policies. One insurer describes the output of the AI models to the FT as “too big a black box.”
The story reminds us that the industry has good reason to be afraid. Google’s AI Overview falsely accused a solar energy company of legal trouble, leading to a $110 million trial in March. Air Canada got stuck in granting a discount last year chatbot invented. And fraudsters used a digitally cloned version of a senior executive to steal last year $25 million from London design agency Arup during a video call that seemed completely real.
What really scares insurers isn’t one huge payout; it is the systemic risk of thousands of simultaneous claims when a widely used AI model steps into them. As one Aon executive put it, insurers can absorb a $400 million loss for one company. What they can’t handle is an AI accident that causes up to 10,000 losses at once.
