The Justice Department’s antitrust division is sharpening its focus on the role artificial intelligence and algorithmic pricing tools can play in facilitating illegal collusion and disguising conduct that would otherwise amount to price-fixing.
Speaking at a recent antitrust conference hosted by the law firm Axinn, Veltrop & Harkrider, Antitrust Division Criminal Deputy Daniel Glad outlined the department’s current thinking on so-called “algorithmic collusion,” arguing that AI systems and pricing software are increasingly becoming mechanisms through which competitors can coordinate prices and exchange sensitive market information.
Glad emphasized that the department does not view AI-driven collusion as fundamentally different from traditional cartel behavior. While algorithms may represent a new technological mechanism, he said the “mechanics of collusion” remain “more conventional than they look on first inspection” and are subject to the same legal standards and evidentiary burdens that apply to traditional antitrust conspiracies.
The remarks provide one of the clearest windows yet into how DOJ intends to pursue antitrust enforcement in an economy increasingly shaped by AI-driven pricing, revenue management platforms and large language models.
Glad argued that algorithmic conduct is not “beyond the reach of criminal antitrust enforcement” when prosecutors can prove that competitors knowingly substituted “shared non-public competitive information” for the independent decision-making required under antitrust law. He said such coordination can occur “through architecture, through information sharing, or through follow-the-algorithm understandings.”
Read more: Maryland Becomes First State to Ban Algorithmic Pricing in Grocery Stores
The comments underscore growing regulatory concern that commercially available software platforms may enable tacit or explicit coordination among competitors without the traditional hallmarks of cartel conduct, such as direct meetings or phone calls between executives.
Glad also suggested that the rise of algorithmic systems may actually strengthen the government’s enforcement capabilities. Software systems generate extensive digital records, including logs, timestamps, queries and data trails that prosecutors can use as evidence. According to Glad, law enforcement’s ability to detect collusion “does not diminish,” but instead “grows” as companies rely more heavily on algorithmic tools.
A key focus of DOJ’s analysis is distinguishing between lawful software arrangements and unlawful “hub-and-spoke” conspiracies in which competitors effectively coordinate through a shared platform provider. Glad said the critical legal question is whether prosecutors can establish a “rim” connecting competitors — in other words, evidence that competing firms knowingly agreed to eliminate competition among themselves.
Glad outlined several factors DOJ may use to determine whether the use of common pricing algorithms crosses the line into criminal conduct.
Among the most significant is whether competitors knowingly rely on each other’s confidential business data. According to Glad, prosecutors will examine whether companies understood that their non-public pricing or market data would shape pricing recommendations provided to rival firms using the same platform.
“If your pricing system depends on your competitors’ confidential inputs to function, you should expect us to ask why that is not anticompetitive coordination,” Glad said.
Glad pointed to the Justice Department’s settlement with rental housing software company RealPage as an example of where the agency believes lawful and unlawful conduct may diverge. Under that settlement, RealPage can continue training models on combined competitor data provided the information is sufficiently aged, aggregated and anonymized.
Beyond proving an agreement, Glad stressed that criminal antitrust enforcement still requires proof of intent. He argued that “intent travels with the human decision to contribute to and rely upon the system,” signaling that DOJ will closely examine how AI tools are marketed, implemented and used inside companies.
In perhaps the clearest summary of DOJ’s position, Glad rejected the notion that technology changes the underlying legal analysis.
“Software cannot launder collusion,” he said. “When competitors exchange competitive intentions in a hotel suite or through a trade association, it is well settled that that raises antitrust concerns. So too with a text thread or a common algorithm.”
