An algorithm does not make decisions in a vacuum: it relies on data, objectives, and rules designed by someone. This is the principle that clearly emerges from the U.S. antitrust action against RealPage and several major real estate operators. On September 4, the Department of Justice announced a proposed consent decree with Pinnacle Property Management Services as part of proceedings into algorithmic coordination practices in the rental market.
If approved by the court, the settlement would prohibit Pinnacle from using anticompetitive algorithms that generate pricing recommendations based on competitors' commercially sensitive information and would impose limits on sharing such information. Under certain circumstances, an independent monitor would also be appointed.
Software does not neutralize collusion
The case is significant because it tackles one of the central issues in modern antitrust: can conduct that would be problematic if carried out directly between companies become acceptable when mediated through common software? The DOJ's stance is firmly negative when the algorithm incorporates competitors' sensitive data and facilitates anticompetitive coordination.
The takeaway extends beyond the real estate sector. Dynamic pricing and automated decision-making systems are increasingly widespread across transportation, commerce, and services. Authorities will have to distinguish legitimate optimization from architectures that reduce competition.
