Lloyd’s of London Offering New Insurance for Chatbot Errors

A popular topic here on TheAICounsel.net is the evolving role of insurance in AI. As the risks associated with AI become more predictable, insurers are developing new AI-focused policies to cover them. The latest to join this trend is Lloyd’s of London, which recently introduced a new product that insures against the legal fees and damages associated with AI chatbot errors. The Financial Times describes the new policy:

“[The policies] cover the cost of court claims against a company if it is sued by a customer or another third party who has suffered harm because of an AI tool underperforming.”

Large language models (LLMs) are notorious for making mistakes. These tools are known to misinterpret or outright fabricate information. Air Canada even faced a judgment ordering them to honor a discount erroneously offered to a customer by their AI chatbot, even though the discount had no basis in company policy. Of course, there are caveats to this new coverage. The policy only pays out if the policyholder’s chatbot performs below initial expectations. So this coverage will only be useful in cases where the chatbot exceeds the expected margin of error. Additionally, coverage will be selective, and not all chatbots will be insurable. The underwriters expect to exclude systems that are especially prone to error and degradation over time. However, the new offering demonstrates the demand for AI-related insurance coverage and may increase AI adoption among those who have been eyeing AI implementation but have sat it out due to risk.