AI liability insurer Armilla names its underwriting backers for the first time: Chaucer, AXIS Capital and Convex

Armilla AI, an insurer of artificial-intelligence liability risk whose coverage is "underwritten by certain underwriters at Lloyd's," publicly named its underwriting capacity providers for the first time this week. "Carriers Armilla is backed by Chaucer, AXIS Capital, and Convex and distributes through licensed brokers in the US, and Canada," the company's site now states. This carrier disclosure does not appear in the capture retrieved Oct. 14, 2026, or in earlier captures retrieved Oct. 9, Oct. 8, Oct. 6, Sept. 11 and July 30.

The company frames the product as covering a gap left by conventional policies: "They are silent on AI failure modes, and some insurers have started adding AI exclusions. Armilla names those failure modes in a standalone policy, so you can deploy AI with coverage in writing." Armilla sells what it describes as "Affirmative AI liability coverage for companies that build and deploy AI, from predictive models to generative AI and agents," with policy limits of "US$25 million aggregate per insured, and up to US$5 million per model." Covered causes of loss include "Defense costs and fines for investigations into AI underperformance, including under the Colorado AI Act and EU AI Act." The product is marketed to companies in "Financial services, insurance, HR and hiring, telecommunications, retail and e-commerce, professional services, and software."

The policy's own illustrative examples point specifically at agentic failures. One: "A procurement firm's AI agent, scheduled with an order-accuracy Model Metric, falls below it and places client orders with the wrong suppliers. Clients claim their extra costs, and the policy responds to those claims." The company qualifies the examples: "Scenarios are illustrative. Coverage depends on the issued policy wording."