Chaucer and Armilla launch AI liability venture
Chaucer and Armilla have jointly launched Vanguard AI, a new structure designed to underwrite cyber, technology, and AI liability risks. The initiative aims to provide streamlined risk assessment and clearer coverage for emerging tech exposures as AI adoption grows in enterprise operations. The offering targets underwriters seeking better risk analytics and claims teams navigating new types of liability.
- The Vanguard AI venture provides dedicated AI aggregate limits of up to $25 million, separate from the $10 million available for cyber-related incidents. This structure is designed to prevent AI-driven losses from depleting traditional cyber or technology errors and omissions (E&O) coverage. - Armilla AI, a Toronto-based startup founded in 2020, provides the standalone AI liability policy for the venture. The company has raised a total of $6.37 million in funding over four rounds and is backed by investors including Y Combinator, Mistral Venture Partners, and Chaucer itself. - Chaucer, a global specialty reinsurance group and member of the China Re Group, brings its established cyber and technology E&O coverage to the partnership. Its Syndicate 1084 at Lloyd's of London, a key part of the business, underwrites a diverse range of specialty insurance and reinsurance classes. - The joint venture aims to address a growing market need as traditional insurance policies are often not designed to cover emerging AI-related risks. This has led other major insurers like AIG and WR Berkley to seek regulatory approval to limit their liability on claims arising from AI systems. - One of the key features of the Vanguard AI structure is the use of predefined allocation rules to determine how coverage will respond when a loss involves a combination of cyber incidents, technology failures, and AI system behavior. This is intended to reduce ambiguity and post-loss disputes about which policy is triggered. - The standalone AI policy from Armilla is backed by Lloyd's of London and specifically covers losses caused by AI model behavior, such as incorrect outputs or underperformance, where no traditional cyber event has occurred. - The collaboration follows Armilla's participation in the Lloyd's Lab accelerator program in 2023, which helps to test and develop new insurance offerings with the support of experts from the global insurance market.
