Policyholders Should Negotiate to Limit AI Exclusions at Renewal

As renewal season approaches, policyholders should be on the lookout for broad artificial intelligence exclusions across all lines of coverage. In the last year, news of insurance companies inserting increasingly expansive AI exclusions has spread.

Risks associated with AI are becoming increasingly widespread and sophisticated. Social engineering fraudsters can now mimic voices and create realistic deepfake videos that are entirely computer-generated. Professionals using AI tools to render services face increased consequences and scrutiny, such as the lawyers sanctioned for using “fake ChatGPT cases” in briefings submitted to the court.

The insurance industry is acting quickly to cut off these and other AI-related liabilities. From insurance services office forms targeting generative AI to the advent of an “absolute AI” exclusion, insurers have made their concern about AI risks well known. For companies whose business models are already inextricably intertwined with the use of AI tools, the potential for significant gaps in coverage is a real risk.

The trend is familiar. When presented with an emerging risk—from Covid-19 to perfluorinated and polyfluorinated alkyl substances (PFAS) or crypto-currency—insurers initially respond with broad-form exclusions and then, over time, calibrate. Early versions tend to be overinclusive, use undefined terms, and use broad “arising out of” language to bar coverage for claims with even a remote or incidental connection to the subject matter.

With AI now touching virtually all aspects of business, from product development to marketing and customer support, sweeping AI exclusions threaten to eviscerate all types of insurance coverage. Policyholders shouldn’t expect AI exclusions to be limited to cyber and media liability policies and should especially be on the lookout for new AI exclusions in their D&O and E&O policies this year.

Below are practical steps policyholders should take to protect themselves.

How to Prepare

First, negotiate now, not after a loss. The most effective countermeasure to broad exclusions is negotiating their removal before a loss. However, where the removal of AI exclusions isn’t possible, businesses should push for clearer definitions, narrower lead‑in language, and targeted carve‑backs that align with their risk profile. For example, negotiating narrower lead-in language can help policyholders avoid surprise denials based on a tangential connection between AI and the loss.

Failing to revise an exclusion that bars coverage for a claim that is “in any way connected to” the use of AI might lead an insurer to deny coverage where the claim arose out of professional services simply because the insured used AI to develop marketing materials. Engaging coverage counsel during policy placement can mitigate the risk of aggressive denials of coverage predicated on broad AI exclusions.

Second, map your AI footprint before applying. Businesses should expect increasingly granular questions on renewal aimed at understanding the extent to which AI is a part of an insured business model. Policyholders need to know—for all aspects of their business—who is using AI, in what way, and for what purpose. By building and maintaining an inventory functions that rely on AI, businesses can respond to questions accurately and avoid the risk of rescission claims premised on incomplete or imprecise answers.

Third, policyholders unable to stave off the addition of broad AI exclusions shouldn’t give up hope when a denial inevitably arrives. Exclusionary language in an insurance policy must be construed narrowly, and courts will not enforce broad language that defeats the purpose of the policy.

Policies with “absolute AI” exclusions will be fertile ground for coverage disputes centering on whether a policyholder’s reasonable expectation to coverage was defeated by a remote connection to AI. In the case of the professional services hypothetical, an insured facing a denial of coverage for a malpractice claim based on the use of AI in creating marketing materials might overcome even a clearly worded broad AI exclusion if enforcing that exclusion would swallow up all coverage provided by the policy. Especially when it comes to the interpretation of new exclusions, businesses should be wary of aggressive coverage positions and work with coverage counsel to identify and address insurer overreach.

AI exclusions are coming fast and sprawling wide. Businesses shouldn’t wait for a claim to find that core protections were hollowed out at renewal. Act now to align your coverage with your risk profile and when faced with a denial of coverage predicated on broad AI-exclusionary language, don’t take no for an answer.

This article does not necessarily reflect the opinion of Bloomberg Industry Group, Inc., the publisher of Bloomberg Law, Bloomberg Tax, and Bloomberg Government, or its owners.

Author Information

Lynda A. Bennett is a partner at Lowenstein Sandler and chair of its Insurance Recovery group.

Alexander B. Corson is an associate in the Insurance Recovery group at Lowenstein Sandler.

Write for Us: Author Guidelines