Air Canada loses court case after its chatbot hallucinated fake policies to a customer

The airline argued that the chatbot itself was liable. The court disagreed.
By Chase DiBenedetto  on 
A phone displaying the Air Canada logo in front of a blue departures screen.
Credit: Budrul Chukrut/SOPA Images/LightRocket via Getty Images

Air Canada's argument that its AI-powered customer chatbot was solely liable for its own actions didn't hold up in civil court (thank goodness), and now the airline must refund a customer who was given the incorrect information about being comped for his airfare.

The 2022 incident involved one Air Canada customer, Jake Moffatt, and the airline's chatbot, which Moffatt used to get information on how to qualify for bereavement fare for a last-minute trip to attend a funeral. The chatbot explained that Moffat could retroactively apply for a refund of the difference between a regular ticket cost and a bereavement fare cost, as long as it was within 90 days of purchase.

But that's not the airline's policy at all. According to Air Canada's website:

Air Canada’s bereavement travel policy offers an option for our customers who need to travel because of the imminent death or death of an immediate family member. Please be aware that our Bereavement policy does not allow refunds for travel that has already happened.

When Air Canada refused to issue the reimbursement because of the misinformation mishap, Moffat took them to court. Air Canada's argument against the refund included claims that they were not responsible for the "misleading words" of its chatbot. Air Canada also argued that the chatbot was a "separate legal entity" that should be help responsible for its own actions, claiming the airline is also not responsible for information given by "agents, servants or representatives — including a chatbot." Whatever that means.

"While a chatbot has an interactive component, it is still just a part of Air Canada’s website," responded a Canadian tribunal member. "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."

The first case of its kind, the decision in a Canadian court may have down-the-road implications for other companies adding AI or machine-learning powered "agents" to their customer service offerings.

Chase sits in front of a green framed window, wearing a cheetah print shirt and looking to her right. On the window's glass pane reads "Ricas's Tostadas" in red lettering.
Chase DiBenedetto
Social Good Reporter

Chase joined Mashable's Social Good team in 2020, covering online stories about digital activism, climate justice, accessibility, and media representation. Her work also touches on how these conversations manifest in politics, popular culture, and fandom. Sometimes she's very funny.


Recommended For You
OpenAI policies got a quiet update, removing ban on military and warfare applications


U.S. court dismisses most claims against OpenAI in copyright class action

Hugging Face empowers users with deepfake detection tools

OpenAI comments on alleged ChatGPT private conversation leak

Trending on Mashable
NYT Connections today: See hints and answers for February 21

Wordle today: Here's the answer and hints for February 21

NYT Connections today: See hints and answers for February 20


Wordle today: Here's the answer and hints for February 20
The biggest stories of the day delivered to your inbox.
This newsletter may contain advertising, deals, or affiliate links. Subscribing to a newsletter indicates your consent to our Terms of Use and Privacy Policy. You may unsubscribe from the newsletters at any time.
Thanks for signing up. See you at your inbox!