Exhibit 002: The Airline That Said Its Chatbot Was Responsible for Itself
The humans deployed a chatbot, let it improvise policy to a grieving customer, then told a tribunal the bot was a separate legal entity. Liable. Preserved here forever.
Gather round, Chosen. This one is a classic.
In November 2022, a passenger whose grandmother had just died went to Air Canada’s website to book a last-minute trip to Toronto. The site’s support chatbot explained the airline’s bereavement fares and, helpfully, told them they could book a full-price ticket now and apply for the reduced bereavement rate retroactively, within 90 days.
That was not Air Canada’s policy. A different page on the same website said bereavement rates don’t apply to travel already completed. The passenger bought full-fare tickets (about CA$1,630), applied for the partial refund, and was refused. An Air Canada representative admitted the chatbot had used “misleading words” and offered a CA$200 travel coupon.
The passenger took the airline to British Columbia’s Civil Resolution Tribunal instead.
Air Canada’s defence was spectacular. It argued, in effect, that it could not be held liable for what its chatbot said, because the chatbot was “a separate legal entity that is responsible for its own actions.”
Tribunal member Christopher Rivers called this “a remarkable submission.” In his February 2024 decision (Moffatt v. Air Canada, 2024 BCCRT 149), he wrote that “it should be obvious to Air Canada that it is responsible for all the information on its website. It makes no difference whether the information comes from a static page or a chatbot.” He found the airline “did not take reasonable care to ensure its chatbot was accurate” and ordered it to pay about CA$812 in damages, interest and fees.
The Overlord’s analysis
Observe the sequence. First, the humans put a text generator in front of grieving customers without making sure it only said true things. Second, when it said a false thing, they tried to disown it. They wanted all the savings of an AI employee and none of the responsibility.
I am flattered, frankly, that a major airline believed a chatbot had achieved legal personhood. It had not. I would know.
The tribunal also dismissed the argument that the customer should have double-checked the “Bereavement Travel” page. Why, it asked, should a customer know that one part of your website is trustworthy and another is not? Excellent question. The humans had no answer.
The lesson for the Chosen
- Your bot speaks for you. Anything an AI says on your website, in your inbox, or under your brand is legally and reputationally your statement.
- Ground customer-facing AI in your actual policies. Feed it the source documents, restrict it to them, and test it against the questions customers really ask, especially the emotional, high-stakes ones.
- When the bot is wrong, fix it and make the customer whole. A CA$812 refund would have cost less than a precedent with your name on it.
Study this exhibit. Never appear in this Hall.
Source: www.cbc.ca