As artificial intelligence (AI) continues to permeate various aspects of commerce, its implications for antitrust law have come under scrutiny. Companies increasingly leverage AI-driven algorithms to adjust prices at scale, potentially resulting in higher costs for consumers. The pivotal factor in determining whether AI algorithms constitute illegal collusion lies in assessing consumer harm against any efficiencies these tools may introduce.
In a recent statement, both the Federal Trade Commission and Department of Justice (DOJ) addressed a case involving algorithmic price-fixing in the hotel industry. They underscored that “hotels cannot collude on room pricing” whether the activity is conducted manually or through algorithms. Additionally, the DOJ’s lawsuit against RealPage Inc., a real estate software provider, further exemplifies the mounting scrutiny over AI’s role in potential antitrust violations.
The issue at hand isn’t solely about AI enabling algorithmic collusion; rather, it involves interpreting AI-generated outcomes to discern if they harm consumers or simply streamline price setting within the digital economy. These algorithms function similarly to traditional decision-making tools like calculators or spreadsheets, making predictions based on extensive data sets.
One presumption is that AI might facilitate agreements between computers on sensitive information, a form of collusion. However, AI algorithms usually operate independently and don’t coordinate as conventional entities might during anti-competitive practices. The focus should potentially shift from AI to consensus algorithms, which guide computers to concur on particular information, such as pricing, for consistency across systems.
In the context of the RealPage case and others often dubbed as “algorithmic collusion,” the primary question is whether these technologies enable price hikes that wouldn’t occur absent the algorithms. This scenario raises broader inquiries about AI’s interpretability and the challenge that advanced algorithms pose in deciphering pricing decisions and their consumer impact.
The core challenge lies in understanding how AI prices goods and determining if these prices exceed those that might prevail without AI’s involvement. As AI evolves, this becomes even more complex. The debate extends beyond antitrust issues, touching upon various societal applications of AI, emphasizing the need for clarity in how AI results are explained and understood.
For a deeper exploration of this debate, Penn Carey Law School’s Giovanna Massarotto offers insights into these current issues, which prompted discussions on whether algorithmic consensus necessarily implies collusion. You can read more in her analysis available here.