Wrongful Act

A wrongful act is the defined event that triggers coverage under a claims-made liability policy, typically an actual or alleged act, error, or omission in the insured's covered professional activities. The phrase is a defined term in the policy, and the precise drafting controls which claims the policy is structured to answer. Disputes over whether conduct constitutes a covered wrongful act are among the most common sources of coverage litigation.

Standard professional liability and Errors and Omissions wordings define wrongful act narrowly around human conduct: an act, error, or omission by the insured in rendering or failing to render professional services. That language was drafted long before AI systems generated content autonomously, and it is not obvious that an AI hallucination is an act, error, or omission of the insured rather than of the model.

Generative AI Liability forms expand the definition to close that gap. Typical wording includes any actual or alleged act, error, omission, or output of a scheduled AI system within the definition of wrongful act. That single drafting choice is what makes the policy responsive to claims arising from AI hallucinations, biased recommendations, defamatory generations, or model misbehavior. Without the expansion, a hallucination is not clearly a wrongful act of the insured, and the policy may decline the claim on definitional grounds.

For a Gen AI deployer this means the defined term is one of the most consequential clauses to read at binding. A policy marketed as covering AI risk but defining wrongful act in the older E&O language may leave model-output claims unaddressed; a policy whose wrongful act definition explicitly enumerates AI system output is what converts the marketing into operative coverage.