Why it matters for your company
Most policies bought before 2025 were written before agents and foundation models were common, and they say nothing about AI. A claim involving a model is then argued under the ordinary grant and exclusions, and both sides have room to argue.
Insurers are closing that ambiguity from both directions, with exclusions on one side and affirmative endorsements on the other (Fenwick, 2026-06-15). For a company whose product is AI, a silent policy is better than an exclusion and worse than wording that names what you do.
Related terms
- Affirmative AI coveragePolicy wording or a standalone product that names AI-related losses as covered, instead of leaving them silent or excluding them. It arrives as an endorsement to cyber or E&O, or as a dedicated AI liability policy.
- Generative AI exclusionAn endorsement that removes coverage for claims arising out of generative artificial intelligence. Standard-form versions for general liability, including one for products and completed operations, took effect in January 2026.
- Limited AI exclusionAn exclusion that removes a defined slice of AI-related claims, such as those arising out of generative AI output, while the rest of the policy keeps responding. The argument at claim time is whether the loss falls inside the slice.
- Technology errors and omissions (tech E&O)Professional liability for technology companies: it responds when your product or service fails to perform and a customer suffers a financial loss. It is the anchor line for an AI or software company.