Conluio por inteligência artificial: devemos nos preocupar?

Objective: the paper discusses Artificial Intelligence (AI) collusion and the potential for AI-driven decision-making to enable entities to coordinate in anti-competitive ways that may evade current enforcement mechanisms. Method: through bibliographic and documentary research, the study examines the implications of AI-supported decision-making on market coordination and collusion. It analyzes enforcement cases in the EU, the US and Latin America and identifies enforcement gaps. Additionally, it explores the differences between EU competition law, US antitrust law and the approach adopted by some Latin American jurisdictions in the context of AI-driven collusion. Conclusions: the main concern with AI-supported pricing decisions from the antitrust perspective is that they may facilitate collusive outcomes outside the reach of antitrust enforcement. Given the likelihood that pricing AI will make it easier for companies to adopt parallel behavior, there is a welljustified fear that it will increase this gray area, where legal parallel behavior and tacit collusion become increasingly indistinguishable. It would be prudent for enforcers in Latin America to follow the European and US examples and consider the risk of AI-driven collusion.