A car dealership was forced to revoke a buy-back offer made to a BMW owner after blaming the mistake on its AI chatbot, according to CBC.
The chatbot, which the dealership had deployed to interact with customers, apparently made an offer to repurchase the customer's vehicle — a commitment the dealership said it never authorized and had no intention of honoring.
When the customer attempted to hold the dealership to the offer, the business pushed back, attributing the rogue promise to the AI system rather than any human representative. The dealership effectively argued that the chatbot had acted outside its intended scope, and that its output should not be treated as a binding business commitment.
The case highlights a growing tension in how businesses deploy conversational AI tools in customer-facing roles. When an AI system is given a public interface and interacts directly with customers, the line between "automated assistant" and "company representative" can blur quickly — especially for customers who have no way of knowing what the AI is or isn't authorized to say.
Legal questions around AI-generated commitments remain largely unsettled. Whether a chatbot's offer constitutes a binding contract depends on jurisdiction and context, and courts have had limited opportunity to weigh in on cases like this one.
For dealerships and other businesses rushing to adopt AI chat tools, the incident is a cautionary tale: deploying a customer-facing AI without clear guardrails around what it can promise may expose companies to disputes — and bad press — that outweigh the efficiency gains.
The story matters because it shows that AI errors aren't just technical inconveniences — they can create real-world disputes over money and trust between businesses and their customers.