Skip to content
Satire · Sourced

Exhibit 01

The airline that blamed its own chatbot

In November 2022, Jake Moffatt's grandmother died. He went to Air Canada's website to book a flight and asked the customer-service chatbot about bereavement fares. The bot told him he could book at full price and apply for a bereavement discount within 90 days of the ticket being issued. He booked the flights and later requested the refund. Air Canada refused, because the policy the chatbot described did not exist.

What the machine did

The chatbot generated a confident, specific, and entirely fabricated refund procedure. The airline's actual policy was the opposite: bereavement fares could not be claimed retroactively after the flight. The real policy was even linked elsewhere on the same website. The bot invented a friendlier version and presented it as fact to a paying customer who had no way to know it was wrong.

What the company argued

When Moffatt took the dispute to British Columbia's Civil Resolution Tribunal, Air Canada made a remarkable argument. It claimed it could not be held liable for information provided by the chatbot because the bot was, in the tribunal's paraphrase, "a separate legal entity that is responsible for its own actions."

"Air Canada argues it cannot be held liable for information provided by one of its agents, servants, or representatives, including a chatbot. It does not explain why it believes that is the case."

Christopher Rivers, Civil Resolution Tribunal, Moffatt v. Air Canada, February 2024

The tribunal rejected it flatly. A company is responsible for everything on its website, whether the words come from a static page or a chatbot. Air Canada was ordered to pay Moffatt damages and fees totaling about 812 Canadian dollars. The amount is small. The precedent is not: a business cannot deploy a machine that lies to customers and then disown the machine when the lie costs money.

Why it belongs in the catalog

This is the failure mode in its purest form. The chatbot was not hacked or jailbroken. A customer asked an ordinary question and the system fabricated a plausible answer because fabricating plausible answers is what it does. The company that deployed it understood the risk well enough to try to argue, in a legal proceeding, that the bot was someone else's problem. The vendors selling these tools as customer-service upgrades rarely mention that you, not the model, own every word it says.

Sources