Estimated reading time: 3 minutes
As the risks of AI grow alongside its benefits, many insurers are taking action to exclude AI-related liabilities from policies. But these AI insurance exclusions appear more common in coverage such as directors and officers (D&O) and errors and omissions (E&O) than cyber insurance, where insurers often cover AI liabilities and are adding coverage. These trends make it even more important for insurance buyers to review gaps and conflicts among their different policies.
Policy Holders Should Double-Check Insurance for AI Exclusions
In management liability and professional indemnity sectors, carriers are deploying exclusions to shield against uncertainties like biased algorithms or faulty AI-driven decisions. For instance, “Berkley’s new exclusion, intended for use in the company’s D&O, E&O, and Fiduciary Liability insurance products, purports to broadly exclude coverage for ‘any actual or alleged use, deployment, or development of Artificial Intelligence,’” according to a legal analysis. This “Absolute” AI exclusion, rolled out in 2025, underscores a trend where insurers define AI expansively to encompass machine learning and generative technologies, potentially barring claims tied even tangentially to AI use.
Berkley is not alone.. “…Philadelphia Insurance and Hamilton Select, have already excluded AI-related claims from E&O policies,” notes a specialized AI insurer’s resource.
“‘They err on the side of an extremely broad definition of not only AI itself, but how it’s being utilized,’ Alexandra Bretschneider from Johnson Kendall Johnson told Insurance Business. ‘Any dependence upon any utilization of AI is really where they exclude.’ That language is now showing up in D&O, E&O, employment practices, fiduciary and crime cover. And it’s the scope – not just the presence – of these exclusions that’s raising red flags.”
AI-Enhanced Perils
In contrast to these artificial intelligence exclusions, many cyber insurers are increasing coverage for AI-enhanced perils rather than excluding them. Risks typically covered include AI-fueled cyberattacks such as deepfake scams and sophisticated social engineering frauds. But even here policy buyers need to ensure their AI risk is covered across different lines. For example, AI can create liability beyond that from breaches or other hacks. “While losses from AI driven cyber-attacks are typically covered in cyber policies, the implications of other risks associated with the adaptation of AI – such as model manipulation, data poisoning, liability arising from hallucinations or wrong output as well as IP infringement – are often not explicitly mentioned in insurance wordings. Innovative products like Munich Re´s aiSure™ – covering the performance of AI solutions – will close potential protection gaps,” asserts Munich Re.
Dedicated Policies for AI Risks?
Standalone AI products remain limited, with offerings like Armilla AI addressing legal harms but excluding bodily injury or property damage. “Armilla’s Affirmative AI Coverage is your fail-safe against fast-evolving AI risks,” the company states, focusing on model performance for developers.
As AI litigation mounts—from deepfakes to IP disputes—insurers’ divergent paths underscore a market in flux, with new technology and evolving legal precedents making it harder for buyers and insured alike to come up with strategies for artificial intelligence liabilities. Perhaps there’s an AI that could help with that…
Related Posts
