As the adoption of algorithmic pricing becomes more widespread, it may be easier than ever for firms to cooperate to raise prices. We set out to demonstrate how reinforcement-learning algorithms can be programmed to reach a collusive outcome.
In a world where the definition of prohibited behaviour is anchored to human intent, what are governments likely to do about digital collusion? We explore what new analytical tools might be needed and what future regulation might look like. Read the full report below: