Netherlands fines Uber $966M for AI-driven driver deactivation

The Dutch data protection authority has fined Uber €825 million, about $966 million, over the way it suspended driver accounts using automated fraud detection. The issue was not simply that Uber used AI. It was that drivers could be deactivated without adequate human review. Uber is appealing the decision.

For any business using AI to make decisions about people, this is a clear warning. If a system flags an employee, contractor, customer, supplier, or applicant, someone accountable needs to review the evidence before a high-impact action is taken. A model can spot unusual behaviour quickly. It cannot reliably understand context, weigh exceptions, or explain a decision in a way that holds up with regulators.

The practical rule is simple: let AI propose, let people authorise. Define which decisions AI can make automatically, which need review, who owns that review, and what record you keep. This is the kind of thing we build into an AI command centre, giving leaders visibility over where AI is acting and where human approval is required.