Commercial insurers are moving to add artificial intelligence exclusions to liability policies, and the shift is happening faster than many policyholders expect. Three ISO endorsements targeting generative AI have drawn growing interest from carriers, according to attorneys and insurance experts tracking the filings.
"There has been a major shift in the insurance industry's treatment of AI-related risks, and insurers are moving very quickly to limit this exposure," said Alana McMullin, a partner at Lathrop GPM whose practice focuses on defending clients in complex insurance disputes.
What the new exclusions cover
The three endorsements, published by ISO parent Verisk, take different approaches to carving out AI risk from commercial general liability coverage:
- CG 40 47 excludes coverage for bodily injury, property damage, and personal and advertising injury arising out of generative AI under the Commercial General Liability Coverage Part.
- CG 40 48 excludes personal and advertising injury coverage arising from generative AI under the CGL coverage part.
- CG 35 08 removes bodily injury and property damage coverage for generative AI under the Products/Completed Operations Coverage Part.
Each endorsement includes a definition of generative artificial intelligence.
McMullin said the filings represent what she calls a reaction to the rapid adoption of AI across business operations. "These ISO exclusions were the spark for this AI exclusion boom," she said. "We're in the middle of what I'd call an industry-wide reaction to the explosion of AI."
She expects carriers to adopt some version of the endorsements at upcoming renewals, though she cautioned that it's unclear how broadly they will be applied. "Insurers are still evaluating their options in real time, so we don't know yet what each insurer will do regarding AI exclusions or how broad their exclusions will be, or necessarily what lines of coverage will contain these exclusions," she said. "But if history is any guide, the anticipated rapid adoption of some version of an AI exclusion in most lines is very likely."
Joe Lam, vice president of liability at Verisk, helped write the endorsements. He said interest is high whenever ISO introduces a new tool, and AI has generated more engagement than most. "Everyone acknowledges this is new technology that will create new exposures that were never contemplated before, so they are appreciative of having additional underwriting flexibility to address any type of emerging issue," he said.
He expects the exclusions to add stability to a market that might otherwise pull back from writing AI risk. "Without exclusions to allow underwriters a level of stability to accept the risk, you run into a situation where they might just walk away from the risk," Lam said. "So, exclusions are very essential in the marketplace."
The problem of silent coverage
Today, many policyholders have what McMullin refers to as "silent coverage" for AI risk - traditional liability policies that don't explicitly exclude it, so a claim might be covered. Insurers could respond in either direction: they may adopt broad exclusions or choose to price and underwrite the risk for an additional premium.
"Insurers have plenty of reasons for adopting these broad exclusions for AI risks, including market pressure," McMullin said. "Aggressive exclusions may make their policies less attractive, especially given the pervasive nature of AI in today's business environment. We may also see insurers who take advantage of the uncertainty in the market and choose to coverage and price the risks for an additional premium rather than they are," she said.
Policyholders may ultimately shape how far these exclusions spread. "All these factors will impact who we see actually implement these policy exclusions and how often they implement them and in which lines," McMullin said.
One carrier has already gone further than the ISO forms. Berkshire Hathaway's Berkley unit introduced an absolute AI exclusion last year for directors and officers, errors and omissions, and fiduciary liability policies. It bars coverage for the use or development of AI, including failure to identify AI-generated content from third parties and an insured's breach of any duty related to AI. Berkley didn't respond to a request for comment.
A surge in AI lawsuits
The interest in exclusions is running parallel to rising litigation. A Gallagher study found a 978% increase in AI-related lawsuits from 2021 to 2025 and a 137% increase from 2024 to 2025. Patent infringement made about 12% of cases, copyright infringement about 11%, and personal injury claims including privacy violations about 10%.
"We've seen these lawsuits that have been filed involving AI - many incorporate disclosure issues, violations of policies, governance, and a company's use of AI and its disclosures to shareholders," McMullin said.
She anticipates that coverage disputes will follow, especially in the E&O and D&O space.
"We think that the E&O and D&O space could be where the coverage lines will expand the AI exclusion more aggressively, put this exclusion into the practice, and potentially have to be litigated when claims are denied," she said.
Why this matters for insurance professionals
As a claims adjuster, underwriter, or broker, you need to track which carriers are filing these exclusions - not just for your own book of business but for the conversations you're having with policyholders before renewal. McMullin said it's difficult to predict where AI insurance litigation might be in months or years, since there is no bellwether case and developments come almost daily.
"The vacancy needs urgent attention," she said. "Policyholders need to understand their potential for AI risk and use other risk-mitigation avenues to close potential coverage gaps," she said.
For those who handle risk for employers or advise clients, that means read the endorsement carefully if a carrier asks you to sign one. Some exclusions go beyond traditional business risks and could leave gaps where you thought coverage still existed - a gap that could be expensive if a claim lands on your desk.
For claims professionals, expect more coverage disputes as these exclusions and the AI litigation wave collide. At a minimum, know what AI for Insurance means for your own roles and questions in underwriting, and direct policyholders with legal questions toward AI for Legal resources.
Your membership also unlocks:
