AI is too risky to insure, say people whose job is insuring risk

Core Insights - Major insurers are seeking permission to exclude AI-related liabilities from corporate policies due to the perceived risks associated with AI outputs being "too much of a black box" [1][2] - The industry has experienced significant incidents involving AI, such as a $110 million lawsuit triggered by Google's AI Overview and a $25 million fraud case involving a digitally cloned executive [1] - Insurers are particularly concerned about systemic risks that could arise from widespread AI model failures, which could lead to thousands of simultaneous claims [2] Group 1: Insurer Actions - Insurers like Great American, Chubb, and W. R. Berkley are requesting regulatory permission to exclude AI-related liabilities from their policies [1] - AIG has clarified that it is not currently seeking to implement these exclusions [1] Group 2: Industry Concerns - The fear among insurers is not just about large individual payouts but the potential for systemic risks that could result in numerous claims at once [2] - An Aon executive highlighted that while insurers can manage a $400 million loss to a single company, they struggle with scenarios where an AI mishap leads to 10,000 losses simultaneously [2]