Insurance claims across multiple U.S. states are seeing carriers begin to incorporate artificial intelligence-related exclusions and policy updates to reflect changes in ISO formats and the growing use of generative AI.
This is one of those slow changes that suddenly become apparent once regulators start filing paperwork.
In Idaho, National Union Fire Insurance Company of Pittsburgh, Pennsylvania, introduced new authorizations for hospice, home health, and related organization programs.
Updates span war, human trafficking, and generative AI. The definition currently includes the heading of generated AI, and this form does not include coverage for personal injury, property damage, personal injury, or advertising injury related to the use of the technology.
AIG, National Union’s parent company, said the application relies on the format of the ISO standard. The company also said it had no plans to implement the exemption, suggesting the language was introduced by default rather than intentionally.
National Union has filed a similar package in Illinois. Local regulators pushed back, asking what real-world scenarios the generative AI exemption was intended to capture.
The news agency responded that AI applications range from basic generation tools to robotic workers, and the possibilities are only increasing.
This approval was positioned as a future underwriting vehicle rather than an immediate change in demand. The application also notes that Illinois has already approved an ISO form with nearly identical language.
Berkeley Insurance has filed similar charges in Connecticut regarding private business liability and criminal offences.
Great American has updated its commercial umbrella and excess forms in Washington to eliminate liability arising from generative AI.
According to our analysts, we are seeing an early outline of the trend. Insurers are currently building optional AI exclusions to help them pivot when loss patterns start to form.
The regulator approved all applications reviewed, suggesting that the regulator views the exemption as consistent with an ISO update rather than a sudden reduction. For buyers, the message is that AI-related risks are entering the underwriting vocabulary, even if the market has not yet decided how aggressively it will use these tools.
