HARMFUL SIGNALS: CARTEL PROHIBITION AND OLIGOPOLY THEORY IN THE AGE OF MACHINE LEARNING

Abstract

Abstract The traditional legal approach for distinguishing between illicit collusion and legitimate oligopoly conduct is to rely on criteria that relate to the means and form of how rivals interact, such as elements of “practical cooperation”, or on the finding of an anticompetitive intent. These criteria ultimately refer to the inner sphere of natural persons and its emanations in communicative acts. Some authors therefore conclude that the cartel prohibition of Article 101 Treaty on the Functioning of the European Union (TFEU) or Section 1 of the U.S. Sherman Act is unable to capture collusion if it is achieved by autonomously acting computers relying on machine learning capabilities. It is instead suggested here to define collusion as parallel informational signals, which achieve a supracompetitive equilibrium, and to use the consumer welfare standard as a proxy for distinguishing between illicit collusion and legitimate oligopoly conduct. This approach is not tantamount to the idea of prohibiting tacit collusion as such. Rather, it is to check singular elements of communication, that is, “informational signals”, within an existing oligopolistic setting for their propensity to create consumer harm. This approach can help to close potential regulatory gaps currently associated with the surge of algorithmic pricing.