In a speech on May 14, 2026, Daniel Glad, the Antitrust Division’s Acting Deputy Assistant Attorney General for Criminal Enforcement, publicly announced that companies using shared algorithmic pricing tools risk not just civil lawsuits but criminal prosecution. The implications of this change in position are significant. Defendants facing prosecution face potential prison sentences for individual executives and substantial fines for corporations.1 For that reason, the Division has historically reserved criminal prosecution for what is considered a “hard core” violation, such as price-fixing, where the judicial precedent is clear that the challenged conduct almost always has anticompetitive effects and lacks any procompetitive efficiency justification.
Companies that use AI to facilitate pricing decisions may now face a similar risk. Although private plaintiffs and civil enforcers have challenged algorithmic pricing in many cases over recent years, the Division had never before articulated a framework for its treatment of these arrangements as criminal offenses. Companies that rely on third-party pricing algorithms or share competitively sensitive data with AI-powered tools now face a fundamentally different antitrust risk calculus.
The Division’s announced criminal theory rests on a hub-and-spoke conspiracy framework, which is the same theory that civil plaintiffs have pursued in the wave of algorithmic pricing litigation over the past several years. Under this theory, competitors (the spokes) each feed competitively sensitive data into a common pricing algorithm (the hub), and the resulting coordination forms the “rim” of a horizontal conspiracy. In Glad’s view, when competitors provide proprietary data to a shared algorithm knowing that the algorithm will use their data to set or influence each other’s prices, they have entered into a horizontal agreement - the essential element of a price-fixing conspiracy. What distinguishes the Division’s posture from prior civil cases is the consequence: Glad contends that when prosecutors can prove such agreements beyond a reasonable doubt, criminal sanctions are warranted.
Factors relevant in assessing use of algorithmic pricing
Glad was careful to note that using an algorithm to set prices is not itself illegal. Companies cross the line, in the Division’s view, when certain factors are present. Based on Glad’s remarks, the Division is likely to consider criminal prosecution when:
- Competitors share proprietary, competitively sensitive information, such as non-public pricing, cost, or supply data, with a common algorithmic platform;
- Participants know (or are willfully blind to the fact) that their data will be used to inform competitors’ pricing, bidding, or other competitive decisions;
- The information exchanged is of the type that would be unlawful to share directly between competitors with or without an algorithmic intermediary; and
- Compliance measures are absent or merely “ceremonial.” The Division will scrutinize whether companies conducted genuine, substantive antitrust assessments before deploying algorithmic tools.
DOJ stakes out hard-core position in face of unsettled legal landscape
Despite the Division’s posture that criminal charges related to the use of pricing algorithms are nothing new, the legal landscape remains deeply unsettled, including in cases where the factors Glad discussed are present. Courts are actively grappling with a threshold question: whether algorithmic pricing arrangements should be evaluated under the per se standard (which treats certain horizontal restraints as inherently unlawful) or under the more permissive rule of reason (which weighs procompetitive justifications). Some courts have held that algorithmic price coordination qualifies as a traditional horizontal restraint subject to the per se rule, while others have concluded that the novel combination of horizontal and vertical elements in a hub-and-spoke algorithmic model warrants rule of reason analysis. No clear judicial consensus has emerged. This leaves companies in a gray area: the Division insists that not all algorithmic pricing is an antitrust violation yet offers little concrete guidance as to when use of an algorithm crosses the line into criminal conduct.
Despite this legal uncertainty, this may be a repeat of the playbook the Division used to begin criminally prosecuting agreements among employers not to solicit or hire one another’s employees. After investigating certain non-solicit and no-poach agreements as potential civil violations in the early 2000s against a patchwork of legal precedent, the Division issued guidance in 2016 declaring that it would consider criminal indictments for such agreements between companies. What followed was a decade of criminal prosecutions that resulted primarily in repeated losses for the Division, but not before companies and individuals spent significantly on their defense.
As of today, the appropriate standard of review for algorithmic pricing remains an open question in many jurisdictions, presenting significant challenges to the Division’s ability to pursue criminal charges. Moreover, proving the knowledge element - that each participant knowingly entered a horizontal agreement through the algorithm - may be particularly difficult when companies rely on third-party vendors and may not fully understand how the algorithm uses or disseminates their data and who else may be using the same software, which may change over time without prior users’ knowledge. The application of criminal law to the use of pricing algorithms has not been tested in the courts. A 2015 case that involved pricing algorithms (although with substantially different facts than the recent wave of algorithmic pricing cases) ended in a plea deal.2 And while the Division explored criminal charges over the use of algorithmic pricing among housing managers, it ultimately chose to bring civil charges rather than criminal charges.3
A&O Shearman’s antitrust team has deep experience counseling companies on algorithmic pricing risk, criminal antitrust exposure, and compliance program design.
Footnotes:
- Daniel W. Glad, Acting Deputy Assistant Att’y Gen. for Crim. Enf’t, U.S. Dep’t of Justice, Antitrust Div., Remarks at the Antitrust West Coast Conference: Old Crime, New Code (May 14, 2026).
- Plea Agreement, United States v. Topkins, No. 3:15-cr-00201-WHO (N.D. Cal. Apr. 30, 2015).
- Complaint, United States v. RealPage, Inc., No. 1:24-cv-00710 (M.D.N.C. Aug. 23, 2024).
