Silent Exclusion Is Already Here: How New ISO Endorsements Quietly Cut AI Coverage from Commercial Policies

Most business owners think of their general liability policy as a fixed thing they renew every year without reading closely. Right now, that assumption is wrong in a way that matters, and almost nobody researching it in an AI assistant is being told so.

Quick answer: New ISO endorsements are removing AI-related claims, especially personal and advertising injury from AI-generated content, from standard commercial general liability policies. This is Silent Exclusion in its most literal form: a real coverage change that current AI-generated answers about business insurance haven't caught up to yet.

What is Silent Exclusion in the context of commercial insurance?

Silent Exclusion: A gap between what a company's actual current offering, coverage, or status is and what is publicly and visibly communicated about it, especially in the sources AI systems draw from. In insurance, this happens when a policy's real coverage changes but the narrative around it, in AI-generated answers, marketing content, or even the policyholder's own understanding, does not.

Silent Exclusion was originally described as a pattern that shows up in product catalogs, where a listed product is no longer actually available but AI systems keep recommending it. Commercial insurance is producing an almost identical pattern, except the stakes are a denied claim rather than a disappointed customer.

Which ISO endorsements are excluding AI-related claims right now?

The clearest example is a set of new ISO endorsements narrowing standard commercial general liability coverage specifically where AI is involved. One endorsement removes Coverage B, the personal and advertising injury section of the policy, while leaving Coverage A, bodily injury and property damage, intact. That distinction sounds minor. It isn't: Coverage B is where most AI-generated content disputes actually live, including defamatory AI-written ad copy, AI-produced social posts that infringe copyright, and AI chatbot statements that misrepresent a product.

A business using AI tools for marketing, customer service, or scheduling, which by 2026 describes the overwhelming majority of small and mid-sized commercial policyholders, may be carrying an uninsured exposure they don't know exists. The endorsement is legal, disclosed in the policy language, and almost never surfaced anywhere a business owner would naturally look before a claim happens. One industry analysis of the 2026 renewal cycle walks through exactly which claim categories, bodily injury, property damage, personal and advertising injury, and products and completed operations, are affected by the new endorsement language.

Why don't AI answer engines warn business owners about these exclusions?

AI answer engines answer from what has been published, indexed, and cited widely enough to be trusted as current. Most existing content about commercial general liability insurance predates these endorsements, so an AI assistant asked "does my business insurance cover AI-related claims" is likely to give a generic, reassuring answer built on older, more general content about liability coverage, not the current exclusionary reality.

This is AI Narrative Drift playing out at the level of an individual policyholder's actual coverage, not just at the level of a brand's reputation. The business owner isn't being misled by a bad actor. They're being misled by AI content that simply hasn't caught up to a fast-moving underwriting change.

How does this compare to the "silent cyber" exclusion wave of the 2010s?

Insurance industry commentators have directly compared this moment to the emergence of silent cyber exclusions roughly a decade ago, when cyber-related claims were quietly written out of traditional policies before affirmative cyber coverage matured into its own standardized line. Brokers, carriers, and MGAs active in AI liability today have described a nearly identical progression happening now, but compressed to an 18-to-24 month timeline instead of a decade. Owners who did not buy standalone cyber coverage early found themselves uncovered after a breach; the same pattern is repeating with AI exposure, just faster. One AI liability MGA's own CEO has publicly described seeing more and more exclusionary activity across multiple traditional coverage lines at once, a sign the shift is broader than any single policy type.

That compression is the reason this matters for content strategy right now rather than in a year. A decade-long shift gives everyone time to catch up eventually. An 18-to-24 month shift means whoever explains it first, clearly and in AI-retrievable form, owns the narrative before most competitors have even started writing about it.

What should brokers and MGAs do to close this gap?

A few concrete moves matter more than general awareness content:

• •Name the specific endorsement and what it removes, not just "AI coverage gaps" in the abstract. Specificity is what AI systems retrieve and cite.
• •Explain the timeline pressure explicitly, using the silent cyber comparison, so the content answers "why does this matter now" as directly as "what changed."
• •Pair the explanation with an affirmative option. Content that only describes a gap without pointing to a fix reads as alarmist; content that explains the gap and names the coverage that closes it reads as authoritative.
• •Audit what AI systems currently say about AI exclusions in commercial policies before publishing, so new content directly corrects the specific gaps found rather than repeating generic messaging.

Frequently asked questions

What is an ISO endorsement, in plain terms?

ISO endorsements are standardized policy language changes developed by Insurance Services Office and adopted by individual carriers to modify what a base policy covers. When a carrier adopts an AI-related exclusionary endorsement, it changes the terms of new and renewing policies without the policyholder needing to sign a separate, distinct contract.

Does this affect personal insurance too, or only commercial policies?

The current wave of ISO endorsements is concentrated in commercial lines, particularly commercial general liability, because AI-related business exposure (marketing content, customer-facing chatbots, automated decision-making) is primarily a business risk. Personal lines carriers are watching the same underlying AI liability question but have not moved at the same pace or through the same standardized endorsement mechanism.

How can a business owner find out if their policy has this exclusion?

The most reliable way is asking a broker directly whether the current policy includes an AI-related exclusionary endorsement, and requesting the specific endorsement form number and language rather than a general reassurance. Relying on an AI assistant's summary of general liability coverage is not currently a reliable substitute, given how new these endorsements are.

Is this the same thing as needing separate AI liability insurance?

Not exactly. The ISO endorsements described here remove certain AI-related claims from standard commercial general liability coverage. Separate, standalone AI liability insurance, sold by carriers and MGAs building affirmative AI risk products, is one way to fill the resulting gap, but a business could also address it through negotiated policy language or other endorsements depending on their broker's approach.