British insurer Lloyd’s of London has unveiled a new insurance policy that covers losses caused by failures or errors in artificial intelligence systems, including generative models like chatbots. About this informs Financial Times.
The policy was developed by Y Combinator-backed startup Armilla and is underwritten through the Lloyd’s network of insurers. The document provides for the coverage of legal costs and compensation in cases where AI activities lead to financial losses — for example, due to the provision of erroneous advice or the dissemination of inaccurate information. Armilla CEO Karthik Ramakrishnan clarified that the insurance applies only in cases of a significant drop in the effectiveness of AI compared to the declared parameters.
The introduction of such a product is a response to the increasing number of incidents related to the use of AI. In particular, in 2024, Air Canada paid compensation to a customer due to a chatbot error, and Virgin Money publicly apologized for an incorrect response related to an ambiguous interpretation of the word “virgin”.
Previously, such risks were partially covered by professional liability insurance policies with a limit of $25,000 with a total limit of $5 million. The new insurance product significantly expands the boundaries of such coverage.
