Recently, a number of prominent insurance companies have submitted applications to U.S. regulatory bodies, aiming to exclude liabilities associated with artificial intelligence (AI) from their corporate insurance policies. This move underscores the industry's profound apprehensions regarding the risks posed by AI. Insurers highlight the unpredictability of AI model outputs, which complicates the process of risk assessment. As the adoption of AI technologies accelerates across various sectors, insurers are encountering a multitude of challenges. These include instances such as erroneous accusations made by Google AI, problems encountered with Air Canada's chatbot, and fraudulent activities involving digital cloning. Each of these incidents has sparked significant compensation claims and legal battles.

Insurers are particularly alarmed by the prospect of a widespread malfunction in commonly used AI models, which could potentially result in tens of thousands of simultaneous claims, posing a systemic risk to the industry. Consequently, the task of accurately evaluating AI risks and devising appropriate insurance policies has emerged as a pivotal concern for insurance companies.