The Chatbot Is Not a Separate Entity

Jake Moffatt asked Air Canada's website chatbot about bereavement fares after his grandmother died. The chatbot told him he could book a ticket and apply for the discount afterward. That was not the airline's policy, and when he applied, Air Canada refused.

Air Canada argued, as the tribunal summarized it, that the chatbot was a separate legal entity responsible for its own actions. In February 2024, a British Columbia tribunal rejected that argument. It found negligent misrepresentation and reasoned that a company is responsible for what its website says, whether the words come from a static page or a chatbot. The award came to about C$812 including interest and fees.

The money was small. The principle is not. A fluent sentence that a customer reasonably relied on belongs to the company that put it in front of them.

Notice what the airline's argument tried to do. It tried to separate the sign from the party who deployed it, so that a statement could exist with no one answerable for what it meant. Organizations are tempted to make that move every time an AI system speaks on their behalf.

It did not work in this case, and it is a poor design principle anyway.

Who in your company signs for what your AI tells customers?