More clarity required on contentious AMS algorithm

The Labor Market Assistance System (AMAS) was supposed to be introduced nationwide at the beginning of 2021, but was stopped in August 2020 by the data protection authority due to concerns over its admissibility. This decision was contested and eventually overturned by the Federal Administrative Court. However, a recent decision made by the Administrative Court has led to further clarification being needed regarding the use of personal data in the algorithm used by AMS.

The main question under scrutiny is whether the digital tool intended to determine labor market prospects would significantly influence decisions made by AMS personnel. This has been under review for almost three years, with concerns raised about profiling and the potential impact on job seekers’ employment opportunities.

The Labor Market Assistance System aimed to divide unemployed people into three categories based on their labor market opportunities using an algorithm, with a goal of allocating funding measures more efficiently and providing support to those with medium prospects. However, it was ultimately up to responsible advisors, such as those tasked with deciding whether someone should receive expensive skilled worker training or not, to make final decisions about unemployment support.

According to a recent decision by the Administrative Court, the algorithm is considered to be in the “significant public interest,” which is a prerequisite for justifying the use of personal data. However, this ruling also confirmed that profiling exists within AMS systems. Whether or not this profiling is admissible will depend on how much influence it has on decisions made by AMS employees when assigning job seekers to different categories.

As a result of this new procedure being necessary for clarification, it remains unclear when and in what form the program could be used moving forward. The AMS is currently reviewing the decision in detail before determining its next steps.

By Editor

Leave a Reply