Air Canada Held Liable for Its Chatbot's Answer

By Satwik · May 20, 2026

In February 2024, a small-claims tribunal in British Columbia ruled that Air Canada must honor a bereavement-fare policy that its website chatbot had described inaccurately. A passenger, Jake Moffatt, had asked the bot about bereavement discounts after a death in the family. The chatbot told him he could book at full fare and apply for the discount retroactively within 90 days. That was wrong; the airline's actual policy did not allow retroactive claims.

Air Canada argued, remarkably, that the chatbot was "a separate legal entity that is responsible for its own actions." The tribunal member rejected this, finding the airline responsible for all information on its website whether it came from a static page or a chatbot, and ordered it to pay damages for negligent misrepresentation.

Why it matters: this is a landmark because it settled, at least in one jurisdiction, that an organization cannot disclaim responsibility for its AI's outputs. The chatbot is not a third party; it is the company speaking. A hallucinated or outdated answer becomes a commitment the business may have to honor.

The defensive lesson bridges security and governance. Ground customer-facing assistants in authoritative, current sources rather than free generation, constrain them to verified answers on high-stakes topics, log interactions for auditability, and recognize that hallucination is not merely a quality problem but a legal and financial exposure. Accuracy controls on a support bot are a form of risk management, not a nicety.