Why are insurers ending silent artificial intelligence coverage?

Over time, risks are expected to integrate into mainstream insurance lines as claims data improves.

Artificial intelligence is changing risk profiles across industries, prompting insurers to move away from “silent AI” coverage under existing policies and towards explicit policy wording, according to an WTW insight.

Currently, AI-related risks are often covered implicitly under cyber, liability, or professional indemnity policies, the “Insurance the AI age” insight by Dr. Anat Lior and Sonal Madhok said. 

This mirrors early cyber insurance, when losses were absorbed under traditional policies before dedicated cyber products existed. 

However, such coverage creates uncertainty, as gaps can arise when AI losses do not clearly fit existing definitions.

Insurers are now introducing AI-specific endorsements or exclusions, signalling a shift towards policies that explicitly address AI. 

Some standalone AI products have also appeared, particularly for small and medium enterprises, whilst larger tech firms often self-insure. 

Over time, AI risks are expected to integrate into mainstream insurance lines as claims data improves.

The evolution parallels cyber insurance: insurers are tightening terms around autonomous decisions and algorithmic errors, making policy reviews at renewal increasingly important. 

Most AI risks can still be mapped to traditional policies, but limits remain—for example, cyber policies may not cover a company’s own data loss, and general liability excludes pure financial loss.

Underwriting practices are adapting, with more detailed questions on AI governance, human oversight, and controls. 

Insurers favour “human-in-the-loop” AI for high-impact decisions, whilst regulatory changes, such as the EU AI Act, are likely to influence liability exposure. Insurers also increasingly act as risk partners, requiring policyholders to implement safety measures to maintain coverage.

Lior said clearer policy language, stronger governance, and improved underwriting data are expected to bring greater certainty, allowing insurance to support safer AI adoption across industries.