Does Your Business Insurance Actually Cover AI Mistakes?

If your business runs an AI chatbot or agent that talks to real customers, do not assume your insurance already covers it if something goes wrong. Starting this year, insurers got new paperwork that lets them write AI claims out of a standard policy, and some already have. The only way to know where you stand is to check your actual policy, not guess.

What Actually Changed With AI and Insurance This Year

In 2026 the insurance industry rolled out three new exclusion forms a carrier can attach to a commercial general liability policy: CG 40 47, CG 40 48, and CG 35 08. No insurer has to use them. But carriers have been asking state regulators for permission to add them, and a lot of that permission has already been granted.

Once one of these sits on your policy, a claim tied to your business's AI can get denied, even if the exact same claim would have been paid out if a person had caused the same harm.

What Do These Exclusions Actually Block?

CG 40 47 is the broad one. It can remove coverage on two fronts at once: the physical injury and property damage side of your policy, and what insurers call personal and advertising injury, the part that normally covers defamation, invading someone's privacy, or using words and images that were not yours to use. CG 40 48 only touches that second bucket. CG 35 08 targets a narrower slice: harm caused by a finished product or a completed job.

For a business running a chatbot or an agent, the personal and advertising injury piece is the one that actually matters. That is the coverage that would normally answer for your AI saying something false about a person, repeating something it shouldn't have, or borrowing content it had no right to use. If your policy carries CG 40 47 or CG 40 48, that answer is gone.

Is This Really a Small Business Problem, or Just a Big Company One?

This is not just a big-company problem. About three out of four small and mid size businesses already use AI in some part of the business, and roughly nine out of ten plan to soon.

“it shall not be a defense, and the defendant may not assert, that the artificial intelligence autonomously caused the harm”

That line is from Section 1714.46(b) of the California Civil Code, added by Assembly Bill 316, which took effect January 1, 2026. It means if your business deployed the AI, you own what it did, whether or not a person typed the actual words that caused the problem.

These exclusions get attached to the same kind of commercial general liability policy almost every small business already carries, not some special big-company policy. If you put a chatbot on your website or an agent answering your DMs, you are exactly who this was written for.

How Do You Find Out If Your Own Policy Already Excludes This?

Call whoever sold you the policy and ask one direct question: does this policy carry an AI exclusion endorsement, and if so, which form number. Most brokers can pull that up in a couple minutes once you ask. Do not accept “we haven't had any issues” as an answer. These forms are new enough that plenty of brokers have not been asked about them yet.

What If You Do Not Have This Coverage?

You have two moves, and you can do both. One is buying coverage built for this. HSB, part of Munich Re, launched a standalone AI liability product for small and mid size businesses in 2026 that covers defense costs and settlements tied to a business's AI use. Ask your broker if something similar is available to you.

The other move costs nothing and matters either way. Cut down how often your AI actually gets something wrong in the first place. That means a real person checking what a customer facing agent is allowed to say before it goes live instead of trusting the default setup, and understanding why a chatbot invents an answer instead of hoping it does not happen again. The more autonomy you hand an AI system, the more that human check matters, not less.

If you want to see what a trained agent actually looks like before you put one in front of your own customers, demo.ajmarketingresults.com runs a live one you can try.

---

Common Questions About AI and Business Insurance

Does my current general liability policy already exclude AI?

Only if your insurer added one of the new exclusion endorsements, CG 40 47, CG 40 48, or CG 35 08, and that is optional for them, not automatic. Ask your broker for the form number and read it yourself rather than assuming either way.

Does this apply to a simple FAQ chatbot, or only advanced AI agents?

The exclusions are usually written broadly enough to cover any system that generates text, recommendations, or responses, not just the advanced kind. A basic FAQ bot answering customers in your voice fits the same definition as a more complex agent.

Do I need to buy separate AI liability insurance?

Not automatically. It depends on what your current policy already excludes, so check first. If you find a real gap, a standalone product like the one HSB launched in 2026, or an added endorsement from your existing carrier, can close it.

Can I still be held responsible if the AI made the mistake on its own?

Yes, at least in California. Assembly Bill 316 removed “the AI acted autonomously” as a legal defense starting January 1, 2026. If your business deployed it, responsibility does not transfer to the software.

What can I do right now if I do not have this coverage yet?

Ask your broker the direct question, and in the meantime keep a person reviewing what your customer facing AI is allowed to say before it goes live. That is the same oversight the new rules already assume you have.