Artificial Intelligence and Collusion (joint with G. Calzolari, S. Pastorello and V. Denicolo)

Abstract:

AI algorithms are increasingly replacing human decision making in real marketplaces. To inform the debate on potential consequences, we run experiments with AI agents powered by reinforcement learning in controlled environments (computer simulations). In particular, we study interaction in the context of a workhorse oligopoly model: price competition with Logit demand and constant marginal costs. We show that independent Q-learners interacting repeatedly with no previous knowledge consistently learn to charge supra-competitive prices. They do so by successfully coordinating on classic collusive strategies. That is schemes providing incentives to cooperate through off-path, incentive compatible and thus credible punishments. We show that this finding is robust to asymmetries in cost or demand structure and to changes in the number of players, product differentiation and demand level.