Do I need special “AI insurance”?

Usually not. Most small businesses are better served by confirming that their existing general liability, professional liability, cyber, crime and employment practices policies do not exclude the AI uses they actually have. A few insurance companies offer affirmative AI wording, and some are adding exclusions, so the wording on your own policy is the thing to read.

Does general liability cover AI mistakes?

Only for the kinds of harm general liability covers, which are bodily injury, property damage and personal and advertising injury, and only if the policy does not carry a generative-AI exclusion. A wrong answer that costs a customer money is normally a professional liability question, not a general liability one.

Is using ChatGPT at work a liability?

It can be, in three ways: what you publish from it, what you type into it and what you decide with it. Publishing unchecked output has led to sanctions for lawyers; typing private data into a public tool can breach your own promises and some laws; and using it to decide about people can trigger hiring-bias rules. A written rule on which of the three your staff may do answers most of it.

Does cyber insurance cover AI?

Cyber policies commonly respond to data breaches and privacy claims whether or not AI was involved, but the part of cyber that pays for a tricked wire transfer is usually a separate, smaller coverage. Ask whether AI-enabled social engineering and a fake voice or video are within it.

What if the AI vendor says it indemnifies us?

Read what it indemnifies. Some vendors promise to defend a customer against a claim that the output infringes copyright, with conditions, and say nothing about wrong answers or biased decisions. An indemnity is only as good as the company behind it and the wording around it.

Does this apply to a one-person business?

Yes. The cases above did not turn on size. A sole proprietor who lets a chatbot quote prices, or who has an AI assistant answer the phone, is making the same statements and recording the same calls as a larger business.

A cafe counter at closing time with a tablet showing a customer chat window beside a printed insurance policy

Using AI in your business: what are the risks, and what does insurance cover?

If your business uses AI, you are responsible for what it says, decides and sends, exactly as if an employee had done it. The six ways it goes wrong are a chatbot or answer that is wrong, a hiring or other decision tool that discriminates, an AI-made image or text that copies someone else’s work, a recorded call or chat that breaks privacy law, a fake voice or video that tricks you into sending money, and private data typed into a public tool. Insurance answers by type of loss rather than by the word AI, and since January 2026 some general liability policies carry an exclusion for generative AI, so check the endorsements list on your own policy.

Call/Text(626) 344-2158

/ Start here

Tell us the situation.

One licensed human replies the same business day — not five agents, not an auto-responder. If the policy you have is already the right one, we will tell you.

We use this to quote and service your insurance, and we do not sell it or pass it to lead networks. Privacy policy.

/ Don’t want to read all this?

Send us what you have. We’ll tell you if it’s right.

Your current policy paperwork, a renewal notice, or nothing at all. A licensed agent reads it and replies the same business day.

Have it reviewed Call/Text(626) 344-2158

The one rule: the business answers for the tool

Every case in this area ends the same way. An airline was held to what its website chatbot told a customer. Two lawyers were fined for a brief that cited cases ChatGPT had invented. A tutoring company paid to settle claims that its application software automatically rejected older applicants. In none of them did “the software did it” work as a defense, because the business chose to use the tool and put its output in front of a customer, a court or an applicant.

That is also how to think about your own use. If an AI tool writes your menu description, answers your phone, drafts your estimate or ranks your applicants, treat the result as something a new employee produced on day one: useful, fast, and in need of a check before it goes out.

The six ways it goes wrong, and what each one looks like

A wrong answer. A website chatbot tells a customer a price, a refund rule, a warranty or a coverage that is not true, and the customer relies on it. The Air Canada case is the clearest example, and the page on chatbot liability covers it in full.

A biased decision. An AI tool that screens applicants, ranks tenants or scores customers can treat a protected group worse without anyone intending it. The EEOC’s iTutorGroup settlement is the case an employer should remember, and several states and New York City now regulate hiring tools by law.

Copied or invented content. Image and text tools can reproduce something that belongs to somebody else, and they can state false things with confidence. This is where media liability and professional liability come in. Our guides on AI-generated content cover how insurance companies underwrite it.

A recorded call or chat. AI receptionists, call-answering tools and website chat widgets often record, transcribe and send conversations to an outside vendor. Lawsuits have argued that doing so without telling the caller breaks state wiretapping and call-recording laws, and the FCC has ruled that AI-generated voices on outbound calls need the same consent as a recorded message.

A fake voice or video. Scammers can now copy an owner’s voice from a few seconds of audio or fake a whole video call, then ask for a wire. This is a crime and cyber question, not a liability one, and the page on deepfake fraud covers where coverage is thin.

Private data typed into a public tool. A customer list, a patient detail or a contract pasted into a free chatbot may be kept or used by the vendor. Whether that is a breach depends on the tool’s terms and your obligations, and the HIPAA and ChatGPT guides cover the health case.

Regulators are already watching

The FTC announced Operation AI Comply in September 2024, five enforcement actions against businesses that used AI to deceive, and its chair said there is no AI exemption from the laws on the books. The practical reading for a small business is that an advertising claim about AI (“our AI lawyer,” “AI-verified”) is held to the same truth standard as any other claim.

State law is moving faster than federal law. New York City and Illinois regulate AI in hiring now; Colorado has rewritten its AI law and pushed the date out; Utah and California have rules about telling people they are talking to a bot. The pages on hiring and on chatbot liability list what to check.

Which insurance answers which AI problem

There is no single “AI insurance” a small business buys. The loss decides the policy, and the question to ask for each is whether the wording leaves AI out.

In January 2026 ISO, the body whose standard forms many insurance companies use, published three optional endorsements that let a general liability policy exclude claims arising out of generative AI. They are optional, so your policy may not carry one, and the declarations or endorsements list will say. The dedicated guide explains each form.

AI problem by policy

The usual home for each loss. Whether your policy actually responds depends on its wording and endorsements.

What went wrong	Policy that usually answers	Where it can fail
Chatbot or AI answer was wrong and a customer lost money	Professional liability or errors and omissions (E&O)	Policy limited to named professional services; an AI exclusion; a claim for a promise or refund rather than a mistake
AI screening tool discriminated in hiring	Employment practices liability (EPLI)	An absolute AI exclusion; wage-and-hour or fines left out; the vendor’s own policy does not protect you
AI image or copy infringed someone’s work	Media liability	Software-copyright and training-data claims; a generative-AI exclusion on general liability
Recorded call or chat sparked a privacy suit	Cyber liability (privacy and regulatory), sometimes media	Exclusions for wiretapping and call-recording statutes are common; check for them by name
Fake voice or video led to a wire	Crime policy or a social engineering endorsement on cyber	Voluntary parting exclusion; a low sublimit; a callback-verification condition
AI-written advice or product injured someone	General liability, or E&O with contingent bodily injury	A generative-AI exclusion; a professional services exclusion
Customer data pasted into a public AI tool	Cyber liability	Excluded if the policy treats the insured’s own disclosure as not a breach; check the definition

Chatbot or AI answer was wrong and a customer lost money
Policy that usually answers
Professional liability or errors and omissions (E&O)
Where it can fail
Policy limited to named professional services; an AI exclusion; a claim for a promise or refund rather than a mistake
AI screening tool discriminated in hiring
Policy that usually answers
Employment practices liability (EPLI)
Where it can fail
An absolute AI exclusion; wage-and-hour or fines left out; the vendor’s own policy does not protect you
AI image or copy infringed someone’s work
Policy that usually answers
Media liability
Where it can fail
Software-copyright and training-data claims; a generative-AI exclusion on general liability
Recorded call or chat sparked a privacy suit
Policy that usually answers
Cyber liability (privacy and regulatory), sometimes media
Where it can fail
Exclusions for wiretapping and call-recording statutes are common; check for them by name
Fake voice or video led to a wire
Policy that usually answers
Crime policy or a social engineering endorsement on cyber
Where it can fail
Voluntary parting exclusion; a low sublimit; a callback-verification condition
AI-written advice or product injured someone
Policy that usually answers
General liability, or E&O with contingent bodily injury
Where it can fail
A generative-AI exclusion; a professional services exclusion
Customer data pasted into a public AI tool
Policy that usually answers
Cyber liability
Where it can fail
Excluded if the policy treats the insured’s own disclosure as not a breach; check the definition

This is a map of where to look, not a promise that any policy pays.

The detail

In Moffatt v. Air Canada (2024 BCCRT 149, decided February 14, 2024) a British Columbia tribunal held the airline responsible for wrong refund information its website chatbot gave a customer, rejected the argument that the chatbot was a separate entity answerable for itself, and awarded about $812 Canadian in total.
American Bar Association, Business Law Today, “BC Tribunal Confirms Companies Remain Liable for Information Provided by AI Chatbot” (February 2024)
In Mata v. Avianca a federal judge in New York fined two lawyers and their firm $5,000 on June 22, 2023 after they filed a brief citing cases ChatGPT had invented, writing that existing rules give lawyers a gatekeeping role to ensure the accuracy of their filings.
LawNext, “Court Imposes Sanctions on Lawyers Who Filed Bogus Cases After Relying on ChatGPT for Legal Research” (June 2023)
In September 2023 the EEOC announced that iTutorGroup would pay $365,000 to settle a suit alleging it programmed its tutor application software to automatically reject women aged 55 or older and men aged 60 or older, rejecting more than 200 qualified U.S. applicants.
U.S. Equal Employment Opportunity Commission, press release, “iTutorGroup to Pay $365,000 to Settle EEOC Discriminatory Hiring Suit”
In the Arup case reported in May 2024, a finance employee in Hong Kong joined a video call in which the company’s chief financial officer and colleagues were all deepfakes, and made 15 transfers totaling about HK$200 million (about US$25.6 million) to five bank accounts; the employee had first suspected a phishing email and was reassured by the call.
CNN Business, “British engineering giant Arup revealed as $25 million deepfake scam victim” (May 16, 2024)
On September 25, 2024 the FTC announced Operation AI Comply, five enforcement actions against businesses using AI to deceive; its chair said there is no AI exemption from the laws on the books, and one company, DoNotPay, agreed to pay $193,000 and stop claiming its “AI lawyer” could replace professional services without proof.
Federal Trade Commission, press release, “FTC Announces Crackdown on Deceptive AI Claims and Schemes” (September 25, 2024)
On August 11, 2025 a federal court in the Northern District of California let a class action proceed against the maker of an AI phone-ordering assistant, Taylor v. ConverseNow, alleging its recording of a customer’s call to a pizza restaurant violated California’s wiretapping and call-recording laws; it is an allegation that survived a motion to dismiss, not a finding of liability.
Wilson Sonsini, “U.S. Federal Court Allows CIPA Class Action Against AI Customer Service Provider to Proceed”
ISO introduced three generative-AI exclusions for commercial general liability in January 2026: CG 40 47 (bodily injury, property damage and personal and advertising injury), CG 40 48 (personal and advertising injury only) and CG 35 08 (products and completed operations); insurers choose whether to attach them.
Gallagher, “ISO introduces generative AI exclusion in commercial general liability policies”
A Gallagher study reported in July 2026 found AI-related lawsuits up 978% from 2021 to 2025, with patent infringement, copyright and personal injury or privacy the largest claim types.
Claims Journal, “Insurer interest in AI exclusions growing as risk becomes omnipresent” (July 20, 2026)
On February 8, 2024 the FCC ruled unanimously that AI-generated voices count as “artificial” voices under the Telephone Consumer Protection Act, so a call using one needs the recipient’s prior consent; the ruling clarified existing rules rather than banning the technology.
Cooley, “FCC: AI-Generated Robocalls Illegal Under the TCPA” (February 15, 2024)

What this means for you

List what AI touches before you list what to insure. Write down every place a tool talks to a customer, decides about a person, records a conversation or moves money. Each line points at one row of the table above.
Read the endorsements list for the word “artificial.” Look on the general liability, the professional liability and the EPLI for CG 40 47, CG 40 48, CG 35 08 or any form that names artificial intelligence, generative AI or a named tool. If one is there, ask what it takes away.
Ask your agent for a written answer on the two or three AI uses you actually have. “Does this policy cover a claim that our website chatbot gave a customer wrong information?” gets a better answer than “do I have AI coverage?”
Put a person in the loop where money or a person’s future moves. A human check on quotes, refunds, hiring decisions and wire requests costs little and is the thing every case above lacked.
Tell people when it is a machine. A short notice that a chat is automated, that a call is recorded and that an applicant is screened by software is cheap, and some states require it.

Go deeper on any of these

Need this handled?

Reading about it and having it handled are different jobs, and the second one is ours. We place the accounts other brokers decline — claims history, a cancellation, an unusual operation, a business nobody wants to underwrite twice — and we will tell you when the policy you already have is the right one. Tell us the situation and a licensed human replies the same business day.