A Dutch court has found that Uber failed to provide two drivers with information about automated account flags that led to reviews of their accounts. The ruling keeps pressure on the company to explain significant algorithmic decisions, while also recognizing limits on how much detail can be disclosed about anti-fraud systems.
The dispute concerns data access and algorithmic transparency under European Union rules. The court’s decision does not determine whether the drivers were correctly classified as potential fraudsters.
What the court found
The Amsterdam District Court ruled in favor of two drivers who argued that Uber had not explained the automated flags that triggered account reviews. It found that Uber had provided no information about those exclusively automated flags, leaving the company in ongoing breach of EU transparency requirements.
Uber also failed to persuade the court to cap daily fines of €4,000 for continued non-compliance. Those fines had reached €584,000. The case puts the practical consequences of data access rights into focus: a company may face continuing penalties when it does not provide information required by a court.
A third driver’s case had a different outcome. The court considered Uber’s explanation clear and sufficient for the time being, even though the driver wanted more detail about where the company drew the line between normal activity and suspected fraud.
How the flag was explained
For that driver, Uber said an automated rule assessed three factors: the number of cancelled rides for which the driver received a cancellation fee, the number of rides performed, and the ratio between cancelled and performed rides over a given period.
The court’s translated ruling said the system signalled potential cancellation fee fraud because the driver performed a disproportionate number of rides in a short period for which he received a cancellation fee. The driver argued that this explanation was too brief to be meaningful without knowing the threshold Uber used.
The interim relief judge sided with Uber on this point. Providing the extra details, the court accepted, could make it easy to commit fraud just below the relevant ratio. The court therefore found the explanation adequate for now, even as the other two drivers prevailed on their claims.
Transparency and anti-fraud controls
The European Union’s General Data Protection Regulation gives individuals protections concerning solely automated decisions with legal or significant effects. It also provides a right to information about algorithmic decision-making, including meaningful information about the logic involved, the decision’s significance, and its envisaged consequences for the person concerned.
The central distinction in this dispute is between an automated flag and the human review that follows it. Uber has said its Trust and Safety Teams review accounts flagged by its systems. The litigation focuses on whether drivers are entitled to information about the automated signal that prompted that review, rather than whether a human team later assessed the account.
Uber has argued that revealing full details of its anti-fraud systems would undermine their effectiveness and has invoked commercial secrets. An earlier Dutch appeals court ruling in litigation involving Uber and Ola found that platforms could not rely on trade secrets exemptions to deny drivers access to data about these kinds of AI-powered decisions.
A continuing legal dispute
The latest ruling adds to long-running litigation in the Netherlands over what platforms must disclose when workers request information about algorithmic management. The cases are testing how to balance a driver’s right to understand a consequential automated decision against a platform’s concern that detailed disclosure could help people evade fraud checks.
The drivers are supported by Worker Info Exchange and the App Drivers & Couriers union. Anton Ekker of Ekker law, which represents the drivers, said they had been fighting for information rights on automated deactivations for several years and described the company’s refusal to comply with the earlier order as objectionable.
Uber said the cases involved three drivers who lost access to their accounts a number of years ago in specific circumstances. The company said its human teams reviewed the accounts after systems detected potentially fraudulent behaviour, and that human review is standard practice.
For now, the court has drawn different disclosure boundaries across the three cases: Uber must address its failure to explain flags to two drivers, while the explanation given to the third driver is sufficient at this stage. The broader question of whether any driver was correctly identified as a fraudster remains unresolved.