Data-driven methods, such as cartel screenings, offer competition authorities new and versatile tools for detecting competition law infringements. To achieve actual practical relevance, however, it is necessary that the outputs of such screenings satisfy certain evidentiary standards. In an analysis carried out within the realm of the DataComp project, Franziska Guggi and Vicky Robertson shed light on this question from perspective of Austrian competition law proceedings. They argue that cartel screenings can be applied throughout different stages of the proceedings: initiating investigations by the Federal Competition Authority, producing reasonable suspicion for conducting a dawn raid, and, last but not least, serving as a basis for the Cartel Court’s decision. To ensure that data is reliable, companies may also be requested to share their internal data. For now, cartel screenings cannot come up with bullet-proof evidence for competition law infringements. It is therefore necessary to keep in mind that an alternative explanation for the screening result is always possible. Last but not least, they stress the importance of explainable AI for competition law proceedings. The paper was published in the latest volume of the Austrian commercial law journal ecolex and can be accessed here.
Source: https://legacy-rdb.manz.at/document/rdb.tso.LIecolex20231136