When Zack Giacomelli, a 31-year-old funeral director in Toronto, decided to sell his 2021 BMW X3 xDrive 30i back to the dealership where he purchased it three years earlier, the process started promisingly. His vehicle needed expensive repairs, so he submitted an online inquiry to BMW Toronto. Soon after, he received a sympathetic text from "Quinn," an eager representative ready to help. 

After exchanging details about the car—which was already at the dealership for service—Quinn made a firm offer: $27,162.79. The amount perfectly covered Giacomelli's remaining loan balance, leaving him with no money owing and no cash back. Excited, Giacomelli countered with $28,500. Quinn responded that it "sounds reasonable" and suggested locking in the deal at 3:30 that day. Giacomelli felt heard and optimistic. 

Then came the shock. Moments later, a human sales consultant called to revoke the offer. Quinn, it turned out, was not a person but an artificial intelligence chatbot. The dealership claimed the AI had misinterpreted data—specifically, confusing the loan payoff amount with a buyback valuation—and proposed a new offer of about $20,000, roughly $7,000 less. 

"I was shocked. I was astounded. Like my jaw was on the floor," Giacomelli told CBC News. He expressed frustration that businesses deploying AI to replace human roles should stand by its outputs. "If they're going to be replacing their employees' jobs with AI, then they need to be honoring what that AI says." 

The incident echoes a 2024 Canadian case where Air Canada was held liable for incorrect information provided by its chatbot, reinforcing that companies are responsible for AI actions as they would be for employee errors. Lawyer Tanya Walker noted that AI can form binding contracts, and undisclosed bot interactions raise fairness issues. 

After media scrutiny from CBC, BMW Toronto reinstated the original $27,162.79 offer. Sales manager Scott Shadbolt said the dealership wanted to "do right" by the customer and acknowledged the AI's misunderstanding stemmed from human input errors. The dealer plans changes, including clearer disclosure when customers interact with AI and restricting bots from independent negotiations. 

This episode underscores growing pains in the auto industry's AI adoption. While chatbots promise efficiency in customer service and lead handling, they risk eroding trust when errors occur without transparency. As Statistics Canada reports rising AI use among businesses (12% in Q2 2025), dealers must balance innovation with accountability to avoid legal and reputational pitfalls. For Giacomelli, relief came after the reversal, but the story serves as a cautionary tale: in the age of AI, a "deal" might not be what it seems.