Algorithmic collusion in international digital markets: challenges to the characterization of collusion and the application of competition law in the age of artificial intelligence

Abstract

This study analyzes algorithmic collusion in international digital markets, focusing on the challenges it poses to the characterization of collusion and the application of competition law in the context of increasing use of artificial intelligence systems. It is based on the hypothesis that traditional criteria, grounded in the demonstration of agreement or meeting of minds, are insufficient to address emerging forms of coordination resulting from autonomous pricing algorithms. The research adopts a qualitative approach, combining literature review and comparative analysis of the European Union, the United States, and Brazil, covering the period from 2015 to 2025. Initially, it examines the economic and legal foundations of collusion, highlighting the transition to digital environments marked by high transparency and real-time monitoring. It then explores the technical mechanisms of automated coordination, emphasizing the role of reinforcement learning in enabling convergence to supracompetitive prices without direct communication among firms. Subsequently, it evaluates the legal treatment of algorithmic collusion in comparative law, identifying greater flexibility in the European model and evidentiary constraints in the U.S. system. Finally, it assesses the challenges faced by Brazilian competition law, indicating that its main limitation lies in the application of traditional categories. The study concludes that algorithmic collusion requires a reconfiguration of analytical and evidentiary standards, shifting the focus from intent to economic effects and system design.