Antitrust Risks to Firms Lurk Inside Some AI Pricing Algorithms

The booming growth of artificial intelligence-based pricing algorithms faces heightened antitrust scrutiny from agencies, state and federal legislators, and private litigants.

Federal Trade Commission Chair Lina Khan last month encouraged enforcers and regulators to be vigilant as AI-pricing algorithms “can facilitate collusive behavior that unfairly inflates prices.” And Congress proposed legislation earlier this year to “prevent anticompetitive conduct through the use of pricing algorithms.”

This growing scrutiny underscores the importance of understanding antitrust laws and policies that are potentially implicated by using AI pricing algorithms.

Sherman Act

Section 1 of the Sherman Antitrust Act prohibits agreements or conspiracies that unreasonably restrain trade. Using AI algorithms may violate Section 1 in two ways.

The first is the traditional handshake conspiracy where competitors agree, explicitly or tacitly, to abide by an algorithm’s pricing decisions. Agreements with horizontal competitors to use pricing algorithms to fix prices, rig bids, or allocate markets are per se unlawful and potentially subject to criminal liability.

The second potential Section 1 violation arises when competitors (the spokes) contract with an AI provider (the hub) to enable competitors to coordinate pricing decisions.

Section 1 allegations connected to AI pricing algorithms have already begun to play out in courts around the country, affecting industries from real estate to hotels. But those claims present challenges for those carrying the burden of proof.

A Las Vegas federal court earlier this year, for example, dismissed a case against a group of hotels where the defendants weren’t required to accept the AI algorithm’s recommendations and where the plaintiffs were unable to “plausibly allege the exchange of confidential information from one of the spokes to the other through the hub’s algorithms.”

Unilateral or mere parallel conduct won’t suffice to prove a Section 1 claim, which means an individual company’s decision to rely on algorithmic pricing isn’t sufficient evidence alone of a conspiracy. Likewise, regulators or plaintiffs must show conduct on behalf of the companies or actors, and AI proposing higher prices may also not be enough to state a Section 1 claim.

Despite these challenges, antitrust agencies have been adamant that conspiracies involving AI pricing algorithms can implicate Section 1 liability.

Section 2 of the Sherman Act prohibits entities from illegally monopolizing or attempting to monopolize a relevant market. While antitrust scholars highlight Section 1 issues posed by AI pricing algorithms, Section 2 concerns are still on the antitrust agencies’ radar.

The Federal Trade Commission cautions that “firms in generative AI markets could take advantage of network effects to maintain a dominant position or concentrate market power,” leading to possible unfair competition tactics by dominant firms.

Additionally, Section 2 price predation concerns could arise should a dominant firm’s algorithms set prices below cost for a sustained period, driving competitors out of the market while subsequently raising prices to recoup losses.

Criminal Antitrust Liability

Severe violations of Section 1 and 2 can expose companies to criminal liability. Although criminal antitrust charges have been rare over the past few decades, agencies have signaled increased willingness to pursue such charges, including in the context of AI pricing algorithms.

The FBI in May, for example, executed a search warrant at corporate landlord Cortland Management’s office as part of the Department of Justice’s investigation of the RealPage alleged AI price fixing conspiracy. In the past several months, the DOJ issued statements of interest in RealPage, another real estate price fixing action, and an Atlantic City hotel price fixing case.

The reported RealPage raid demonstrates the DOJ’s willingness to follow through on its October 2021 announcement to strengthen its response to corporate crime. Potential criminal liability for coordinated use of algorithmic pricing introduces the possibility of prison sentences for individuals, as well as substantial fines for convictions.

FTC Act

Section 5 of the FTC Act protects unfair or deceptive acts affecting commerce. The FTC has been active in exercising its Section 5 authority and stated that Section 5 is broader than the antitrust laws.

Section 5 is also a basis for the FTC’s 2023 lawsuit against Amazon.com, Inc. Given the FTC’s current interest in AI, the FTC might use Section 5 as a mechanism to fight AI-enabled price discrimination.

Merger Clearance

Antitrust agencies are now broadly reviewing deals for a wide range of potentially anticompetitive conduct, examining factors such as interlocking directorates, non-compete agreements, and potentially AI-related concerns in the clearance process.

The FTC and DOJ’s merger guidelines, issued in December 2023, specifically caution that using pricing algorithms could signal an increased risk of coordination between competitors, indicating use of algorithms may be assessed in the antitrust deal clearance process.

Guidance

Given heightened antitrust scrutiny, firms using AI pricing algorithms should take several high-level steps to help comply with US federal antitrust law.

Avoid unlawful agreements. Avoid entering into any agreement with competitors to use AI pricing algorithms as a way to fix prices, rig bids, or allocate markets.

Know your algorithm. Understand how your AI pricing algorithm gathers information, including whether it uses information from competitors.

Document pro-competitive benefits. When implementing AI pricing algorithms, it helps to document the pro-competitive benefits from the process. This may include increasing consumer price transparency, and increasing efficiency and decreasing costs in determining competitive prices.

Include AI in merger due diligence. Understand whether and how the parties involved in the deal are using AI pricing algorithms and whether that use could raise scrutiny in the deal clearance process.

Consult counsel. Legal counsel can clarify the legal risks and framework associated with algorithmic pricing and assist with updating compliance policies and training to reflect best practices.

While AI algorithms promise new means of efficiency for an array of industries, the AI boom is also fair game for antitrust enforcers. Their recent activity shows they are watching and ready to employ a number of enforcement strategies. Companies eager to implement AI into their business should be aware of recent activity and take steps to make sure they comply with antitrust laws.

This article does not necessarily reflect the opinion of Bloomberg Industry Group, Inc., the publisher of Bloomberg Law and Bloomberg Tax, or its owners.

Author Information

Emily Renzelli is counsel at Rule Garza Howley, focusing on antitrust litigation and government investigations for health-care and technology clients.

Whitney Williams is an associate at Rule Garza Howley, focusing on antitrust regulatory clearance and government investigations.

Erica Baum is an associate at Rule Garza Howley, focusing on antitrust litigation and government investigations.

Write for Us: Author Guidelines