The short version
- "The AI did it" is not a defense. Air Canada argued its chatbot was responsible for its own actions; the tribunal rejected that.
- Professionals answer for AI-assisted work. Lawyers who filed fake AI-generated citations were sanctioned, and the duty to check work does not change because a tool drafted it.
- Regulators treat AI claims like any other claims. The FTC has said there is "no AI exemption from the laws on the books."
- E&O is the main protection. It covers claims that your services caused a client financial loss. Confirm it covers services delivered with AI and has no AI exclusion.
The cases
| Case | What happened | Result |
|---|---|---|
| Moffatt v. Air Canada, 2024 BCCRT 149 | Air Canada's website chatbot told a customer they could apply for a bereavement fare after travel. That was wrong. | The British Columbia Civil Resolution Tribunal found negligent misrepresentation and held Air Canada responsible for all information on its website: "It makes no difference whether the information comes from a static page or a chatbot." It awarded about CAD $650 plus interest and fees. |
| Mata v. Avianca, Inc., 678 F. Supp. 3d 443 (S.D.N.Y. 2023) | Lawyers filed a brief citing cases ChatGPT invented. | The court found bad faith and imposed a $5,000 sanction on the lawyers and their firm, and required letters to the client and to the judges falsely named as authors. |
| FTC order against DoNotPay (2025) | A company marketed a "robot lawyer" without support for its claims. | The FTC's final order required $193,000 in monetary relief and barred unsupported AI lawyer claims. |
| FTC v. Rite Aid (2023) | The FTC said Rite Aid's AI facial recognition falsely flagged shoppers as shoplifters. | The settlement bans Rite Aid from using facial recognition for surveillance for five years and requires deletion of data and models derived from it. |
The Air Canada decision is a small-claims ruling in Canada and does not bind U.S. courts, but its reasoning is the one businesses should expect: the company answers for its own website and tools.
Who is exposed
- Businesses with customer-facing chatbots quoting prices, policies, eligibility, or instructions
- Consultants, agencies, and analysts whose reports or recommendations are drafted with AI
- Law, accounting, and financial firms where AI errors become professional negligence claims
- Software companies whose AI features give users wrong outputs they act on
- Anyone advertising what their AI can do, where overstated claims draw regulators
How insurance responds
| Claim | Policy | What to check |
|---|---|---|
| Client says your AI-assisted work caused a financial loss | Professional liability (E&O) or technology E&O | Definition of professional services; any AI exclusion; claims-made dates |
| Customer relies on wrong chatbot information | E&O, sometimes general liability | Whether answering customers is a covered service; AI exclusions |
| Regulator investigates AI marketing claims | Some E&O and D&O policies cover investigation costs | Fines are often uninsurable; defense may be covered |
| AI output is a pure performance failure | Affirmative AI policies such as Armilla's | Availability and limits |
See does cyber insurance cover AI for why cyber usually is not the answer here.
Steps that reduce the risk
- Keep a human in the loop on anything sent to a client or published as advice.
- Limit what the chatbot can promise. Prices, refunds, and eligibility should come from approved sources the bot quotes, not from what it generates.
- Say when customers are talking to AI. Washington and Idaho chatbot disclosure laws starting in 2027 mainly target companion and consumer chatbots and exclude most customer service bots, but disclosing AI is still good practice; see AI laws for Washington and Idaho businesses.
- Read your engagement letters and contracts for AI clauses and limits of liability.
- Confirm your E&O covers AI-assisted services at every renewal.