Your renewal came in looking like last year's: same limits, same carrier, a short list of form numbers on the declarations page. One of those numbers may be new, and it may say that claims arising out of generative artificial intelligence are no longer covered. For a business that now answers customers with a chatbot, drafts its marketing with AI tools, or screens job applicants with software, that one line can matter more than the limit.
What ISO released for 2026
ISO, the organization that writes the standard forms most general liability policies are built on, made three new endorsements available with a January 2026 edition date:
- CG 40 47, Exclusion – Generative Artificial Intelligence. For the Commercial General Liability coverage part (occurrence and claims-made). It excludes bodily injury and property damage (Coverage A) and personal and advertising injury (Coverage B) arising out of generative AI. This is the broad version.
- CG 40 48, Exclusion – Generative Artificial Intelligence (Coverage B). Same coverage part, narrower reach: it removes only personal and advertising injury arising out of generative AI, leaving bodily injury and property damage alone.
- CG 35 08, Exclusion – Generative Artificial Intelligence. For the separate Products/Completed Operations liability coverage part, excluding bodily injury and property damage arising out of generative AI.
All three are optional. ISO doesn't force them onto anyone; each insurer decides whether to file and attach them, on which accounts, in which states. Trade press reported that by mid-2026 insurers had made thousands of generative AI exclusion filings across nearly every state, either the ISO forms or their own versions. A filing doesn't put the form on every policy, so you can't assume your policy has one, and you can't assume it doesn't.
Notice the trigger. "Arising out of" is broad causal language in most courts. The exclusion doesn't ask how much of the work AI did, only whether the claim connects to it.
It isn't only general liability
Insurers are adding their own AI language to other lines, and those forms aren't standardized:
- Professional liability / E&O. AI-specific exclusions are showing up as professionals use AI tools in client work. Some are narrow; some reach any claim involving AI output.
- Directors and officers. Some management liability forms now carry broad AI exclusions reaching claims "based upon, arising out of, or attributable to" AI, which can capture lawsuits alleging the company overstated its AI capabilities to investors.
- Cyber. Wording varies widely. Some forms add AI limitations; others now affirmatively address AI-driven fraud such as deepfaked voice or video used to trick staff into paying.
- Media liability. If your content is AI-assisted, check how the policy treats infringement or defamation in material a model generated.
The direction isn't one-way. Some insurers are also offering affirmative AI coverage, endorsements or standalone policies that expressly cover defined AI exposures. That's the counterweight worth asking about.
Who's actually exposed
This isn't a tech-company problem. The exclusions follow the use of generative AI, not the industry. If you do any of the following, you have exposure:
- A customer-facing chatbot. In *Moffatt v. Air Canada* (2024 BCCRT 149), a British Columbia tribunal held the airline responsible for its website chatbot telling a grieving customer he could claim a bereavement fare after travel, contrary to the airline's own policy. The airline argued the chatbot was responsible for its own statements; the tribunal rejected that. It's a Canadian small-claims decision and binds no U.S. court, but the principle (your bot speaks for you) is the one plaintiffs will argue.
- AI-written marketing. Ad copy, images and product claims generated by AI can infringe someone's copyright or trade dress, or defame a competitor. Those are classic Coverage B claims, which is precisely what CG 40 48 removes.
- AI in professional work. An accountant, consultant, designer or agency that delivers AI-assisted work owns the error if it's wrong. That's an E&O claim, and the E&O form may now have its own AI carve-out.
- AI hiring tools. In *Mobley v. Workday*, pending in federal court in California, an applicant alleges an AI screening tool discriminated by age, race and disability. The court let the case proceed on the theory that a software vendor can be liable as the employer's agent, and in May 2025 preliminarily certified a nationwide age-discrimination collective. There's no finding of liability, and the case is still moving. Employers using these tools should expect discrimination claims to name them too, which is an employment practices liability question.
The gotcha: the exclusion you didn't notice
Most AI exclusions don't arrive with a letter. They arrive as one more form number on the schedule of a policy that otherwise renewed "as expiring." Nobody calls it out, because on paper the limits and carrier didn't change.
You find out at claim time, when the denial letter quotes an endorsement you never read. And because the trigger is "arising out of," a claim that is mostly about something else (a misleading ad, a botched deliverable, a customer who relied on bad information) can be pulled in if AI touched it.
The second trap is assuming your other policies will catch what GL drops. They might, or the E&O, D&O or cyber form may carry its own exclusion with different wording, and the gap lands in between.
What to do at renewal
- Read the forms schedule on every policy's declarations, not just GL. Look for anything mentioning artificial intelligence, generative AI, algorithms or automated systems.
- Ask directly, in writing: does an AI exclusion apply, to which coverage parts, and how does it define AI? A narrow definition and a broad one are very different policies.
- Map your actual AI use. Chatbots, content tools, AI in client deliverables, hiring software. Each points to a different line of coverage.
- Ask what affirmative AI coverage is available if an exclusion is unavoidable, and whether another line picks the exposure up.
- Align your contracts. If you sell AI-enabled services or buy AI tools, check the indemnity and insurance clauses. A contract may require coverage your new exclusion removed, and your vendor's terms may push the risk back to you. Talk to your attorney about the contract language.
Our approach
At Cory Washington & Co., we read the endorsement schedule on every renewal, not just the declarations page, and we flag AI exclusions by coverage part before you bind. We start with how you actually use AI, then check that your general liability, professional liability, cyber, media, EPL and D&O forms fit together instead of leaving the exposure stranded between them.
For technology companies and AI and machine learning companies, where AI is the product, we look at technology E&O built for that risk. For everyone else, the goal is the same: know what's excluded before a claim tells you. If you want background first, see professional liability vs. general liability and cyber insurance for small business.
Renewal coming up, or not sure whether an AI exclusion is already on your policy? Request a policy review and we'll read the forms with you.
Frequently Asked
What is the ISO generative AI exclusion?
Starting with a January 2026 edition, ISO made three optional endorsements available that let insurers exclude generative AI exposures from general liability. CG 40 47 removes coverage for bodily injury, property damage and personal and advertising injury arising out of generative AI; CG 40 48 removes only personal and advertising injury; and CG 35 08 does the bodily injury and property damage version for the separate Products/Completed Operations coverage part. They are optional, so whether one is on your policy depends on your insurer and your renewal.
Does my general liability policy cover claims caused by an AI chatbot?
It depends on the policy wording and whether an AI exclusion is attached. Many chatbot problems, like a bot giving a customer wrong information that costs them money, are financial or professional in nature and may not fit general liability at all. If a generative AI exclusion is endorsed onto your GL, claims arising out of generative AI can be excluded even where coverage would otherwise have applied. Read the forms schedule on your declarations to see what's attached.
Do I need to be a tech company to worry about AI exclusions?
No. The exclusions are written around the use of generative AI, not the type of business. A retailer with a website chatbot, a firm that writes marketing copy with AI tools, a professional who uses AI in client work, or an employer that screens applicants with automated software can all have an AI-related claim. The question is whether your policies still respond to it.
How do I find out if my policy has an AI exclusion?
Look at the forms and endorsements schedule on each policy's declarations page and search for endorsements that mention artificial intelligence, generative AI or automated systems. Check every line, not just general liability: professional liability, D&O, cyber, media and employment practices policies can carry their own AI language. If you can't tell, ask your broker to confirm in writing which coverage parts carry an AI exclusion and what it says.
This article is general information for business owners, not insurance or legal advice, and does not bind or alter coverage. Policy terms, eligibility, and pricing vary by carrier and state — confirm specifics with our licensed team before making decisions.