Emerging risk · 5 min read
Your business is using AI. Your insurance might have quietly stopped covering it.
Most owners have spent two years learning to use AI. Very few have been told what it is quietly doing to their insurance. The wording is starting to shift, and it works the way a coverage change always seems to. Nothing looks different on the surface, right up until a claim.
This has happened before, with cyber
A few years ago, cyber losses were often covered under general policies simply because nothing said they were not. The industry called it silent cyber. Then insurers added clear exclusions, standalone cyber policies matured, and coverage people assumed they had was suddenly spelled out, or spelled away. The same pattern is now playing out with AI. The industry even uses the same phrase for the old state of things: silent AI.
What actually changed
At the start of 2026, the standard commercial liability policy forms used across the United States introduced a specific exclusion for generative AI. It is broad. It reaches bodily injury, property damage, personal and advertising injury, and the products and completed operations side of a policy, when the claim arises out of the use of generative AI. Generative AI is defined widely enough to capture the everyday tools businesses already use to create text, images, audio, video, and code.
Put plainly, when that exclusion is attached at renewal, a claim tied to your use of AI can fall outside a policy that looks otherwise unchanged.
Why a Canadian business should care
These forms are American, so this is not a statement that your Canadian policy already excludes AI today. It is a statement about direction. Canadian commercial wordings tend to follow the US and London markets, and the silent cyber story shows how quickly an implicit coverage becomes an explicit exclusion once the market decides to move. If you carry US exposure, sell across the border, or renew with an insurer that leans on these forms, the change can reach you sooner than you would expect. The point is not to panic. It is to look before the market looks for you.
Where AI is probably already in your business
This is not only a technology-company concern. If your business uses AI to draft content or marketing, answer customer questions, screen job applicants, generate code, summarize documents, or support decisions, you are already in the zone these exclusions are written about. The exposure is not the tool. It is what happens when the output causes harm, and which policy is supposed to respond when it does.
Three questions worth asking now
- Does my current policy already exclude AI, or is it likely to at my next renewal?
- If my use of AI leads to a claim, which policy is actually supposed to respond: liability, cyber, professional liability, or none of them?
- Is anyone actually checking this for my business, or am I assuming I am covered?
Frequently asked questions
Does my commercial policy still cover AI?
In the US, standard general liability forms added a generative AI exclusion effective January 2026. Canadian wordings often follow the US and London markets, so the safest answer is to check your specific policy rather than assume.
What is silent AI?
It describes AI risks being implicitly covered only because nothing excludes them, the same way silent cyber worked before explicit cyber exclusions and standalone policies arrived.
What should I do if my business uses AI?
Confirm whether your policy already excludes AI or will at renewal, identify which policy is meant to respond to an AI-related claim, and have a licensed broker review your wording before an incident.
Not sure where your AI exposure sits?
Run your policy through the free Coverage Assessment, or have me read your actual wording and flag where an AI claim would, or would not, be covered.
By J.R. Genua, CCIS, Certified Cyber Insurance Specialist.