Recent ISO endorsements addressing artificial intelligence appear designed to limit or exclude coverage for losses arising from generative AI processes. However, ISO forms represent only part of the insurance marketplace. Many insurers use proprietary forms, manuscript endorsements, or modified policy language that may approach AI-related exposures differently. We will examine several non-ISO examples to identify how carriers are revising definitions, exclusions, and coverage provisions and what those changes could mean when determining whether a claim involving AI is covered, restricted, or excluded altogether.