Protecting Against the Next Wave of AI Product Liability Exposure

By: Michael S. Nelson and Nicholas N. Chan

With the growing shift towards artificial intelligence product liability litigation, policyholders should assess their exposure and coverage risks under their existing policies. As previously discussed, plaintiffs are testing whether consumer-facing AI applications should be treated not as traditional services, but as products that would be subject to claims such as design defect, inadequate warnings, and foreseeable misuse. As courts determine these threshold questions for liability purposes, policyholders would benefit from exploring AI-specific coverage to protect against these risks and should be aware of emerging AI-specific exclusions in the marketplace.

In 2024, we first wrote about coverage for AI risks. By then, insurers had begun to introduce products designed to protect against the particular risks that companies involved with generative AI models might face. Cyber insurance add-on endorsements might purport to expand coverage to address risks such as data poisoning, infringement, and usage and regulatory violations. Nonetheless, policyholders should be mindful that their cyber policies may have exclusions for liabilities assumed by contract.

Technology errors and commissions (Tech E&O) insurance policies provide coverage for third-party claims that allege a wrongful act, error, or omission in the performance of technology services or the failure of a technology product to perform as intended. Tech E&O policies generally provide express coverage for breach of contract claims, but they typically exclude coverage for product liability claims and liability arising out of bodily injury or property damage.

Policyholders should not overlook traditional coverages such as commercial general liability (CGL), directors and officers (D&O), and errors and omissions (E&O). However, the increased scrutiny from underwriters, narrowing of policy terms, and emerging exclusions for AI-related losses signal the need for policyholders to assess their current AI-related exposures under their existing policies and explore specialty coverage to close any potential gaps. In particular, the Insurance Services Office introduced AI-related CGL endorsements last year. For example, ISO advises that CG 40 47 excludes GL coverage for bodily injury, property damage, and personal and advertising injury arising out of generative AI. Form CG 40 48 is somewhat narrower, in that it preserves coverage for AI-related bodily injury and property damage. Even if policyholders have confirmed that their existing policies do not contain AI-specific exclusions, they should be aware that renewals may now contain materially changed language to exclude coverage for AI-specific losses.

Therefore, policyholders, especially those who use, develop, deploy, or integrate AI applications, should take three immediate steps: (1) identify where AI is used, including in products, services, workflows, customer interfaces, marketing, and decision-making; (2) examine existing policies for possible coverage, including the definitions, conditions, and exclusions that could impact AI product liability claims; and (3) explore AI-specific specialty coverage to close identified gaps in existing policies.

The rapid adoption of AI is now an insurance recovery issue.
Policyholders would greatly benefit from evaluating how AI-related claims may fit within existing coverage, scrutinizing any new exclusions at renewal, and considering whether specialized AI coverage is needed to mitigate the next wave of AI product liability litigation.