Antitrust Agency Insights: Developments At The US Antitrust Enforcement Agencies'Second Quarter 2022

• Published date: 06 July 2022
• Subject Matter: nti-trust/Competition Law, Antitrust, EU Competition
• Law Firm: Arnold & Porter
• Author: Ms Sonia Kuester Pfaffenroth and Matthew Tabas

Successfully navigating antitrust agency investigations requires a familiarity with Department of Justice and Federal Trade Commission processes, as well as insight into those agencies and their leaderships' current priorities for enforcement and competition policy. This newsletter will provide periodic updates on both, offering an analytical look at how the antitrust agencies are approaching important competition issues and what current investigations may mean for potential future enforcement. We hope our experience'both inside and outside these agencies'will provide insights that help you make more informed decisions for your business.

Letter from the Editors

Increased Focus on Algorithmic Pricing and Potential for AI Coordination

Antitrust enforcement has been criticized in the past for failing to keep up with technological advancements.¹ One new technological quandary is the use of algorithmic pricing, allowing companies to increasingly price dynamically in near real time in response to competitor price movements, customer demands and other market conditions. The US Department of Justice, Antitrust Division (DOJ) and other observers have raised concerns about the potential impact of this practice'and artificial intelligence (AI) more generally'on competition, including whether it increases the potential risk for price fixing through technological means.²

DOJ has signaled to companies that they should be cognizant of the risk of price fixing when technology is used. At the 21st Annual International Competition Network Conference in Berlin, Germany on May 4, 2022, Assistant Attorney General (AAG) Jonathan Kanter acknowledged the interplay of AI and antitrust. AAG Kanter suggested that, "whether you use a smoke-filled room in a basement or you're using AI and an API, it's still the same thing. It's still collusion."³ AAG Kanter suggested that companies proactively design (and train) algorithms and artificial intelligence programs not to collude, just as they train their employees.⁴ To address these new market dynamics, DOJ is increasing its capacity to pursue investigations and enforcement actions in this area.⁵

It is clear that if two firms (or their employees) expressly agree to use artificial intelligence and/or algorithms as a means of fixing prices, they face antitrust risk just as much as if they agreed to use a more primitive implementation.⁶ The use of algorithms alone, however, should not give rise to antitrust liability. Sherman Act ' 1 prohibits agreements to fix prices for competing products, but does not prohibit mere "parallel pricing" (i.e., competing products being sold at the same or similar prices absent proof of an agreement).⁷ And Sherman Act ' 2 jurisprudence generally protects the right of any company, even a monopolist, to set prices freely.⁸

But what happens if artificial intelligence and algorithms are able to collude among themselves without human intervention? In theory, and validated by some experiments, pricing programs in the same market can "agree" to a common price, and appear to raise prices in a coordinated way, without express instructions from human overseers to do so.⁹ Historically, US enforcers have maintained a consistent position that the legal standard for finding unlawful collusion does not change in the context of pricing algorithms'the independent use of pricing algorithms by two companies, without an agreement to fix prices, should not give rise to liability.¹⁰ Yet, AAG Kanter's suggestion that AI could facilitate the equivalent of a "smoke-filled room," suggests that DOJ may be entertaining the idea that AI is capable of entering into an unlawful "agreement" for purposes of Sherman Act ' 1. And, at the very least, algorithm pricing that is correlated with competitors' pricing could be viewed as a plus factor (combined with other factors), that could lead to an inference of an anticompetitive agreement.

Courts will likely continue to demand proof of conspiracy (i.e., an actual agreement between human beings) in order to prove allegations of price fixing, but AAG Kanter's speech emphasizes that algorithmic pricing and other automated programs may raise scrutiny. DOJ is building up its capability to investigate, and potentially challenge, conduct that may cross the line from unilateral decision making to inappropriate coordination between competitors. Parties using pricing AIs or other similar systems should consider and take steps to mitigate antitrust risk when developing, purchasing and/or supervising those programs. Finally, antitrust compliance training will become even more important for companies using algorithmic pricing programs to ensure that no one'or no system'crosses the line.

• Read Media Coverage
• Read more from our editor Sonia Pfaffenroth

Continued Focus on the Pharmaceutical Supply Chain

During the week of June 13th, the DOJ and Federal Trade Commission (FTC) hosted two events highlighting their focus on the pharmaceutical industry, particularly the role of pharmaceutical benefit managers (PBMs) and the impact of rebating practices in the industry. On June 14th and 15th, the DOJ...