Cowbell unveils AI liability insurance for small businesses that deploy AI agents

Cyber insurance firm Cowbell Cyber Inc. introduced Cowbell AI Liability Insurance, a policy for U.S. small and midsized businesses. This standalone policy covers losses from AI agents, such as those that misread approval statuses, manipulate customer records, or are manipulated by attackers. Coverage also extends to inaccurately generated output, unauthorized usage charges, and "LLMjacking," where attackers use stolen credentials to run AI workloads.

Claims from customers fall under the third-party side, while regulatory proceedings and certain fines are also covered. Cowbell's Chief Commercial Officer, Simon Hughes, emphasized that AI liability insurance now extends beyond situations where humans approve AI actions, highlighting the need for protection when AI agents operate autonomously.

Underwriting relies on Cowbell's three AI-specific factors—observability, autonomy, and governance—to assess a business's AI risk profile. The new policy, slated for launch on January 1, 2027, will be sold through wholesale brokers on a non-admitted basis. Cowbell already offers AI coverage through its Prime One policy, but this new product is designed specifically for smaller businesses.