AI Exclusions in Liability Insurance: What Business Owners Must Know in 2026
Photo by Morthy Jameson on Unsplash
- Major insurers including AIG, W.R. Berkley, and Great American have filed to exclude AI-related claims from standard liability policies.
- Verisk (ISO) released standardized AI exclusion endorsements — CG 40 47 and CG 40 48 — effective January 1, 2026, now available to every U.S. carrier.
- If your business uses AI tools such as chatbots, AI agents, or generative AI, your existing policy coverage may no longer protect you from related claims.
- Standalone AI liability coverage is emerging, with specialists like Armilla AI offering up to $25 million per organization for risks traditional insurers are now walking away from.
What Happened
If your business uses AI tools — a customer service chatbot, an AI-powered hiring assistant, or a generative AI content tool — you may have quietly lost some of your insurance protection without ever receiving a warning letter.
Three of the biggest names in U.S. commercial insurance — AIG, W.R. Berkley, and Great American — have sought regulatory approval from state regulators to offer liability policies that exclude AI-related claims. That includes risks tied to chatbots, AI agents, and other automated systems your business might depend on every single day.
To make it easier for every carrier in the country to follow suit, the insurance industry's standard-setting body, Verisk (also known as ISO), released two new endorsements — CG 40 47 and CG 40 48 — effective January 1, 2026. These forms give all U.S. carriers a plug-and-play option to strip generative AI exposures out of Commercial General Liability (CGL) policies — the standard business liability policy that covers bodily injury, property damage, and personal injury claims.
W.R. Berkley went even further, proposing what it calls an "absolute AI exclusion" — broad language designed to bar any claim "based upon, arising out of, or attributable to" any actual or alleged use, deployment, or development of AI. This sweeping exclusion would span D&O (Directors & Officers liability), E&O (Errors & Omissions, also known as professional liability), EPLI (Employment Practices Liability Insurance), and fiduciary liability lines.
AIG confirmed to the Financial Times that it filed generative AI exclusions but stated it "has no plans to implement them at this time" — treating the filings as optionality for future use. Even so, the fact that filings exist at all signals clearly which direction the entire market is heading.
Photo by Igor Omilaev on Unsplash
Why It Matters for Your Coverage
Think of your business insurance policy like a Swiss Army knife. For years, that knife has quietly included a blade for "AI-related mistakes" — even though neither you nor your insurer ever explicitly discussed it. That hidden blade is what the industry calls "silent AI" exposure (existing policies that may inadvertently cover AI-related claims without using the word AI anywhere in the contract). Now, insurers are pulling it out and saying: "We never agreed to cover that."
This matters enormously for your policy coverage — and for your overall risk assessment as a business owner. Until recently, if an AI tool your company used made a discriminatory hiring decision, generated false information that harmed a customer, or triggered a privacy violation, there was a credible argument that your existing CGL or E&O policy might cover the resulting lawsuit. Those arguments are getting much harder to make in 2026.
The Verisk CG 40 47 endorsement alone covers both Coverage A (bodily injury and property damage) and Coverage B (personal and advertising injury) — the two broadest pillars of a standard CGL policy. When your carrier attaches this form to your policy at renewal, both pillars no longer support AI-related claims. For businesses running AI tools customer-facing, that is a significant hole.
The professional liability market has already experienced what analysts describe as a "structural break" between January 2025 and January 2026 — a period where gradual tightening rapidly became a sudden market bifurcation on AI coverage. In plain English: the market did not slowly drift; it snapped. A careful insurance comparison today will reveal a very different landscape than the one that existed just twelve months ago.
The regulatory environment is accelerating this shift. The NAIC (National Association of Insurance Commissioners) AI Model Bulletin — which sets governance and accountability expectations for how insurers use and respond to AI — has been adopted by 23 states plus Washington D.C. as of April 2026. As more states align with these standards, insurers have even greater incentive to formally define where AI risk sits and who ultimately pays for it.
For small business owners, this is a critical moment to conduct a real insurance comparison across your commercial lines. If you renewed your policy before January 2026 without reading the endorsements carefully, you may not know whether AI exclusions have quietly been added at renewal. A thorough side-by-side insurance comparison — ideally with a licensed broker who specializes in commercial lines — could reveal significant gaps in your protection before a claim ever hits.
The good news is that the market is responding. Armilla AI, the first Lloyd's of London coverholder (an entity authorized to underwrite insurance on behalf of Lloyd's syndicates) dedicated exclusively to AI liability, now offers standalone AI Liability Policies covering up to $25 million per organization. That kind of specialized coverage is becoming a core component of smart risk assessment for any company deploying AI at meaningful scale.
Chasing insurance savings by choosing the cheapest policy at renewal could cost you dearly if those savings come at the expense of policy coverage for your largest emerging exposure. The least expensive policy is not the best policy if it has a gaping hole where your AI risk used to live.
Photo by Steve A Johnson on Unsplash
The AI Angle
The irony is hard to miss: the insurance industry is using AI-driven underwriting tools to decide which AI risks it will no longer cover.
Modern risk assessment in commercial insurance increasingly relies on AI-powered data analysis to flag emerging exposures — and AI liability has lit up every model. Dennis Bertram, head of cyber insurance for Europe at Mosaic Insurance, told the Financial Times plainly: "It's too much of a black box." Insurers cannot predict the frequency or severity of AI-related claims with the actuarial precision (the statistical science of pricing insurance risk) required to offer coverage responsibly at scale.
Rajiv Dattani, co-founder of the Artificial Intelligence Underwriting Company, framed the core problem even more starkly: "Nobody knows who's liable if things go wrong."
This uncertainty is reshaping claims management across the industry. Insurtech platforms are stepping into the gap — Armilla AI was built specifically to underwrite and manage AI liability claims that traditional carriers cannot model. Meanwhile, AI-powered claims management tools are being deployed by carriers to triage and classify incoming AI-related claims even as standard policy language moves to exclude them. The future of claims management in this space will demand expertise that sits at the exact intersection of technology law, AI systems knowledge, and insurance coverage analysis.
What Should You Do? 3 Action Steps
Pull out your most recent CGL, E&O, and D&O policies and look for any new endorsements — particularly any language referencing "artificial intelligence," "generative AI," "automated systems," or "chatbots." If you see CG 40 47 or CG 40 48 attached to your CGL policy, your AI-related exposures under Coverage A and Coverage B are no longer protected under that policy. Do not wait until a claim is filed to make this discovery. An insurance comparison with a broker who specializes in commercial lines can help you understand exactly what you have lost and what alternatives are available to replace it.
Before your policy renews, create a straightforward inventory of every AI tool your business uses — customer service bots, hiring tools, content generators, AI-assisted financial or legal advisory tools, and any third-party software that uses AI under the hood. Each one represents a potential liability exposure that your insurer is now actively working to exclude. This inventory forms the foundation of a credible risk assessment and will help any broker identify the right standalone or supplemental coverage to fill the gaps. With the NAIC AI Model Bulletin now adopted across 23 states plus D.C., documenting your AI governance practices also has growing regulatory compliance value beyond just insurance.
The growing standalone AI liability market — led by specialists like Armilla AI with policy limits up to $25 million per organization — exists precisely because traditional carriers are retreating. Ask your licensed broker about standalone AI liability policies and how they interact with your existing commercial lines. Yes, this adds to your premium. But think of it as the opposite of insurance savings in the short term: a manageable additional cost today versus a completely uncovered multimillion-dollar claim tomorrow. Always consult a licensed insurance agent or broker before making any changes to your coverage structure.
Frequently Asked Questions
Does using AI tools at my business automatically void my existing liability policy coverage in 2026?
Not automatically — but it can create significant coverage gaps depending on your carrier and your specific renewal terms. The Verisk ISO endorsements CG 40 47 and CG 40 48, effective January 1, 2026, give every U.S. carrier a standardized way to exclude AI-related claims from CGL policies. Whether your policy includes these exclusions depends on what your carrier attached at your most recent renewal. Review your policy documents carefully, look for new endorsements, and consult a licensed insurance agent to get a clear picture of what your current policy coverage does and does not include when AI is involved.
What does W.R. Berkley's absolute AI exclusion mean for my business insurance premiums going forward?
W.R. Berkley's proposed "absolute AI exclusion" would remove AI-related claims from D&O, E&O, EPLI, and fiduciary liability policies entirely — lines that cover executives, professionals, and employers respectively. If your carrier adopts similar language, you may need to purchase separate standalone AI liability coverage to maintain equivalent protection, which could affect your total insurance costs. Businesses that do not use AI at all may find this kind of exclusion largely irrelevant. A thorough insurance comparison with a knowledgeable broker can help you determine the net impact on both your premiums and the breadth of your policy coverage at the next renewal.
How does the NAIC AI Model Bulletin affect my company's risk assessment and eligibility for business insurance in 2026?
The NAIC AI Model Bulletin, adopted by 23 states plus Washington D.C. as of April 2026, sets expectations for how insurers govern and account for AI in their own internal operations. For businesses, the significance is indirect but increasingly real: as regulators push insurers to be more rigorous about AI risk internally, insurers are becoming more rigorous in underwriting businesses that use AI externally. Your company's own AI governance practices — how you oversee, document, and audit your AI tools — are becoming a growing part of the risk assessment conversation when insurers decide how and whether to cover you. Strong internal AI governance may ultimately help your insurability.
Is there any liability insurance that still covers AI-related claims for small businesses in 2026?
Yes — but you will likely need to look beyond standard CGL or E&O policies, which are the most affected by the current wave of exclusions. The standalone AI liability market is developing specifically to fill the gap left by traditional carriers stepping back. Armilla AI, operating as a Lloyd's of London coverholder, offers AI Liability Policies covering up to $25 million per organization. Other specialty insurers and MGAs (Managing General Agents — specialized underwriters who focus on niche or emerging risk markets) are also building AI-specific products. Claims management under these policies is handled by teams with dedicated expertise in AI liability, which matters given the complexity involved. Always consult a licensed agent to find the coverage that fits your specific business and AI use case.
Can strong AI risk management practices lead to insurance savings on commercial liability policies in 2026?
Potentially yes, and the trend is moving in that direction. As the AI liability insurance market matures, specialty underwriters are beginning to differentiate businesses based on the quality of their AI governance — how thoroughly they document AI use, audit for bias or errors, and maintain human oversight of automated decisions. Businesses that can demonstrate strong internal risk assessment frameworks around AI may be viewed as lower-risk and could qualify for better pricing or broader policy coverage terms over time. While widespread premium discounts for AI governance are not yet standard, the underlying logic is sound: proactive risk management today positions your business for better insurance savings and stronger coverage options as this market continues to develop. Talk to a licensed insurance professional about how your AI practices might factor into your next renewal conversation.
Disclaimer: This article is for informational purposes only and does not constitute insurance advice. Always consult a licensed insurance agent for personalized guidance.
No comments:
Post a Comment