While the tech world is racing to embrace and integrate AI into business processes, the fact that these software tools have become too risky to be insured has sparked a new debate within the industry. According to a report published by the Financial Times, this technology, which everyone is chasing, now poses unacceptable risks for insurance companies. Insurance giants such as AIG, Great American, and WR Berkley have begun filing applications with US regulators to exclude AI-related liabilities from corporate policies. One insurer in the industry describes the output of AI models as an unpredictable “black box.”

Can AI errors be insured? Billion-dollar fears are growing

The insurance industry’s concerns on this issue are not without merit, as recent events highlight the magnitude of the risk. In March, Google’s AI-powered overview feature inadvertently linked a solar energy company to legal problems, leading to a $110 million lawsuit. Similarly, Air Canada was forced to offer its customers a discount created by a chatbot last year. Fraudsters also used a digital replica of a senior executive to conduct a highly realistic video call to steal $25 million from London-based design firm Arup.

The scenario that truly terrifies insurers isn’t a massive payout to a single company. The real concern is the systemic risk that could arise if a widely used AI model fails, leading to thousands of claims simultaneously. An Aon executive summarizes the situation succinctly: “Insurers might cover a $400 million loss to a single institution. But what’s unmanageable is a chain reaction catastrophic loss triggered by an autonomous AI, causing 10,000 other losses simultaneously.”

While AI technologies make our lives easier, corporate and financial risks appear to be growing. What are your thoughts on this? Will these insurance restrictions imposed by companies on the use of AI slow the technology’s pace of adoption? Don’t forget to share your thoughts with us in the comments.