“According to Air Canada, Moffatt never should have trusted the chatbot and the airline should not be liable for the chatbot’s misleading information because Air Canada essentially argued that “the chatbot is a separate legal entity that is responsible for its own actions,” a court order said.”

Can you imagine the hellscape we’d be living in if precedent went the other way? Companies could just run every unsavory decision through some machine learning system and then wash their hands of it afterwards.

“Oh you were illegally fired? Sorry, that decision came from the Overmind, not from us.”

Well i see a future for very profitable prompt injection.

Ha, these fucking assholes. Glad to see they got forced to honor it. That’s what you get for jumping on bad tech in order to save a few bucks.

Hope other companies face the same liability if they rely on AI chatbots.

Pretty confident if this happened because they hired a new (real) support operator who just didn’t understand the policy, they would have made a concession to the customer and the support person would likely just get more training.

But because it’s a chat bot that they really don’t understand (outside of their IT department), they go to court and shut down a system they likely spent hundreds of thousands of dollars developing.

This type of advanced decision making is why we pay CEOs the big bucks.

It certainly cost them more in lawyers and co, than paying it directly and quietly as gesture of goodwill.