Uber would let its algorithms determine which drivers should be fired, the news recently reported. Three British drivers and one Portuguese brought a case to the court in Amsterdam. Uber had blocked the "driver accounts" of these drivers for fraudulent practices. The question is: Does this case actually involve an algorithm dismissal? And so is the algorithm fear in this case justified? Yes and no.
Yes, because from what I understand the British drivers are employed by Uber. I am not an English law specialist, but then this does smack of a dismissal. But from a Dutch perspective, this is different. Dutch drivers usually work for Uber through a zzp-construction. So then there is no employment relationship. Hence the no.
Yet I think there is more to be said for unwarranted algorithm anxiety. Uber has discovered - apparently - using digital tools that its drivers are engaging in fraudulent practices. Let's face it: is this so much news under the sun? Again, I'm just wondering out loud. Because I know plenty of examples of companies using digital tools to have people's wrongdoings investigated and brought to the surface. When this comes to the surface, immediate dismissal usually follows. And again, I ask myself: is what Uber is doing here that much different?
Drivers' case against Uber has been filed under the AVG. According to the drivers, there is a violation of the prohibition against an automated decision. For the enthusiast: Article 22 AVG. This article reads as follows:
"The data subject has the right not to be subject to a decision based solely on automated processing, including profiling, which produces legal effects concerning him or significantly affects him in any other way."
Please note the word "exclusively" here. Uber will no doubt argue that this is not the case, and there is human intervention in the decision.
Furthermore, there is an important exception, namely, the prohibition does not apply if "necessary for the establishment or performance of a contract between the data subject and a controller."
Therefore, if the court rules that there is "exclusively," I think the substantive hearing will revolve around this exception. I think this is a good development in itself, because if you ask me there is a need for more interpretation and more depth on this point. Especially also in this setting.
My opinion is now clear. It is a nice piece of "framing" by the Dutch lawyer for the drivers, but what is happening here does not concern "dismissal by an algorithm. So we are being unfairly given an algorithm scare. But it is an interesting development that a Dutch judge is looking into the scope of the ban on an automated decision.
To be continued!
More articles from BDO
