BMW dealership reinstated the AI offer after hearing from CBC News
This is somewhat on topic:
A German regional court has ruled that Google is directly liable for false claims in its AI-generated search overviews – (Archived link)
The Regional Court of Munich hit Google with a temporary injunction barring the company from spreading false claims about two Munich-based publishers through its AI-generated search overviews (case no. 26 O 869/26). The court classified Google as a direct infringer because the “AI overview” is its own content, not just a list of search results.
As Canadian businesses rush to adopt artificial intelligence tools, they face a growing risk of customer backlash — even legal action — if those tools make mistakes.
Canadian law has already established that companies can be held liable if AI chatbots dole out bad information. In a 2024 case, Air Canada was forced to honour a fare rebate after its chatbot provided a passenger with incorrect advice about bereavement fares.
The airline argued before the British Columbia Civil Resolution Tribunal that the chatbot was “a separate legal entity that is responsible for its own actions.” But the tribunal disagreed, stating that Air Canada was ultimately responsible.
“Just like an employee may do something wrong and the company’s held responsible, a bot is just like an employee,” said Tanya Walker, a litigation lawyer with Walker Law in Toronto.
“I don’t think companies really realize the magnitude and the power that a bot can have,” she said. “It can enter into a contract on your behalf.”
The airline argued before the British Columbia Civil Resolution Tribunal that the chatbot was “a separate legal entity that is responsible for its own actions.”
That had to be a complete hail mary legal tactic, they couldn’t have possibly thought that was going to fly. As the article says, even a human employee making a mistake doesn’t get them out of liability.
It was worth a shot (from the perspective of an evil, soul-sucking corporation) - had it gone their way, it would have been one hell of a legal precedent.
BMW should have just stayed quiet, bit the bullet, and got IT to fix the problem rather than say “yeah we do not value our workers, we are trying to replace them all with AI.”
Shadbolt said that, due to miscommunication from a human employee, Quinn misinterpreted the amount Giacomelli owed on the car — $27,162.79 — as the amount BMW would pay to buy back the vehicle.
As soon as it is within the context window, there’s always a risk that LLMs just start running with that.
Imagine if the customer said “I’m trying to buy a new BMW that’s $34997”. Then asked the AI what an appropriate sell price would be for his current car including “trade-up incentives” or something like that.
I’m guessing that one could make a killing by getting AI to agree to bad deals.
Feature not a Bug of AI.
