Collusion by Code: What DOJ’s Algorithmic Pricing Enforcement Means for the Energy Sector

Introduction

On May 14, 2026, Acting Deputy Assistant Attorney General for Criminal Enforcement Daniel Glad delivered remarks titled “Old Crime, New Code” at the Antitrust West Coast Conference in San Francisco. The speech represents one of the most detailed public statements to date by a senior DOJ official on the application of criminal antitrust law to algorithmic pricing tools, software-as-a-service (“SaaS”) platforms, and large language models (“LLMs”). For companies operating in the energy sector, where algorithmic tools increasingly drive wholesale power pricing, fuel procurement, capacity bidding, and trading strategies, the implications are significant and immediate.

We believe there is no longer any doubt that the DOJ intends to apply existing criminal antitrust doctrine to algorithmic and AI-driven conduct with the same rigor it has applied to traditional cartel behavior. Energy markets present a particularly acute set of risks: concentrated market structures, repeated competitor interactions in organized wholesale markets, realshared data platforms for capacity and pricing information, and a long history of antitrust scrutiny dating back to the Enron era. Companies that develop, deploy, or rely on algorithmic pricing and procurement tools should act now to audit those tools, update their compliance programs, and prepare for an enforcement environment in which the digital evidence trail is longer, the pool of potential whistleblowers is wider, and the window for self-reporting is narrower than ever before.

This Legal Update summarizes the key themes of Glad’s remarks and provides practical compliance guidance tailored to in-house and outside counsel advising energy companies, utilities, traders, and power market participants.

Key Takeaways for the Energy Sector

Glad’s remarks carry several important implications for energy companies that use or develop algorithmic pricing tools, revenue management software, or AI-driven trading and bidding systems.

First, criminal liability is on the table for algorithmic collusion. Glad stated unequivocally that “software cannot launder collusion” and that the use of algorithms, SaaS platforms, or LLMs to coordinate pricing among competitors does not place such conduct beyond the reach of criminal prosecution under the Sherman Act. For the energy industry, where third-party platforms routinely aggregate generation cost data, fuel price indices, and capacity availability across market participants, this principle has direct application. Where evidence shows that competitors used a software system to replace independent decision-making with shared competitive intelligence, the DOJ will treat that conduct as it always has treated cartel conduct.

Second, the DOJ’s own investigative tools are increasingly effective in the digital environment. Energy markets generate massive digital footprints, including algorithmic bidding records, trading platform logs, and automated dispatch communications, all of which are discoverable and can serve as evidence of coordination.

Third, the Whistleblower Rewards Program has created a “second race” alongside the Leniency Program. Energy companies now face two parallel races: the corporate race to self-report through the Leniency Program and the individual race of insiders to claim a whistleblower reward. In January 2026, the Division announced its first-ever whistleblower reward: one million dollars paid to an individual whose information led to bid-rigging charges in online auctions. The energy trading environment, where personnel routinely have access to algorithmic strategies and competitive data, broadens the pool of potential whistleblowers considerably.

Fourth, the DOJ is actively developing its framework for AI-generated pricing. The DOJ’s posture is that existing doctrine is sufficient and that the classification of conduct does not change because the software is newer. For energy companies deploying AI-powered forecasting and bidding models, this means the same antitrust rules apply regardless of whether prices are set by a human trader or an AI-driven optimization engine.

Fifth, antitrust compliance programs must rigorously address AI and algorithmic tools.

The Hub-and-Spoke Risk in Energy Markets

A significant portion of Glad’s remarks focused on the hub-and-spoke theory of algorithmic antitrust enforcement. He identified RealPage as the “paradigm case,” noting that the Antitrust Division entered a consent judgment requiring the company to limit the granularity of its pricing outputs and submit to a court-appointed monitor. The DOJ action came against the backdrop of already-pending civil litigation challenging RealPage’s software platform, which landlords allegedly used to set rents. According to the government, the tool worked by gathering competitively sensitive, non-public information from competing landlords and feeding it through a shared algorithm that generated pricing recommendations, effectively allowing competitors to align their pricing through a common intermediary rather than through direct communication. The matter was resolved as a civil enforcement action analyzed under the rule of reason, not as a criminal prosecution. The remedy was targeted at the specific mechanics that turn a pricing tool into a vehicle for exchanging competitor information, rather than at algorithmic pricing as a general practice.

For energy companies, the hub-and-spoke paradigm maps directly onto common market structures. Third-party platforms that aggregate real-time generation cost data, fuel procurement prices, or capacity bids from competing generators or marketers can function as the collusive “hub.” When multiple competitors feed proprietary operational data into a shared platform that then generates pricing recommendations, the risk that the platform operates as an information exchange is substantial.

The remedy in RealPage was targeted at the specific mechanics that turn a pricing tool into an information exchange, rather than banning algorithmic pricing generally. Energy companies should take note: the DOJ has signaled that it is not opposed to algorithmic tools as such, but that it will scrutinize tools that pool non-public competitor data and produce coordinated outputs. And while RealPage itself was civil in nature, companies should not read that resolution as a signal that algorithmic conduct sits beyond the reach of criminal enforcement.

AI-Generated Pricing: Emerging Risks for Energy Trading

The final substantive portion of Glad’s remarks focused on the emerging questions surrounding AI-generated pricing. While Glad emphasized that he was not previewing a charging theory, he outlined three questions the Antitrust Division is actively considering.

First, what constitutes the “agreement?” Most AI providers train their models on user inputs by default under their terms of service, which could make such a model “the collusive hub for anticompetitive spokes.” For energy companies using AI-powered trading platforms, the risk is that proprietary trading strategies, pricing data, and capacity information shared with an AI vendor could be incorporated into models serving direct competitors in the same wholesale market.

Second, where does intent lie? The relevant inquiry is what the actor knew and what the actor did with that knowledge, not who typed the code. Energy traders and quantitative analysts who understand that a platform aggregates competitor data could face personal liability for deploying such tools without appropriate safeguards.

And third, does the per se rule apply? The Division’s posture is straightforward: “the classification does not change because the software is newer. The classification tracks the conduct.” Where competitors have agreed to eliminate competition among themselves, the per se rule applies regardless of the technological medium.

Compliance Guidance for Energy Companies

We believe the DOJ has now provided pointed guidance on compliance in the algorithmic era. The Antitrust Division updated its Evaluation of Corporate Compliance Programs in Criminal Antitrust Investigations in November 2024 to include risk-assessment questions directed at AI and algorithmic tools. Those questions ask whether a company’s risk assessment addresses its use of new technologies; whether the company assesses antitrust risk when deploying new tools; whether compliance personnel are involved in that deployment; and what steps the company takes to mitigate risk.

Glad has now warned that AI governance programs focused solely on privacy, cybersecurity, intellectual property, and bias, without addressing antitrust, are incomplete. He stated: “A company cannot say it has a mature AI governance program if no one is asking whether the tool facilitates coordination with competitors.”

In light of these developments, in-house and outside counsel advising energy companies should consider the following practical steps:

Conduct an Algorithmic Tool Inventory and Data-Flow Audit

Energy companies should compile a comprehensive inventory of all pricing, procurement, trading, bidding, and revenue-management tools currently in use across the organization. For each tool, counsel should map the data flows: what data the tool ingests (including whether it incorporates non-public competitor data such as generation costs, fuel procurement data, or capacity information), how the tool processes that data, and where the outputs are directed. This audit should extend to third-party SaaS vendors whose platforms may aggregate data from multiple competing market participants. In energy markets, particular attention should be paid to platforms used in wholesale power auctions, gas trading, and capacity markets where competitors routinely interact through shared infrastructure.

Review Third-Party Vendor Agreements and Terms of Service

Glad’s observation that most AI providers train their models on user inputs by default underscores the importance of reviewing vendor agreements and terms of service for all algorithmic and AI tools. Counsel should examine whether vendor contracts permit the use of a company’s proprietary data, including pricing, capacity, generation, and supply information, to train models that serve competitors. Companies should negotiate contractual protections, such as data-use restrictions, opt-out provisions for model training, data segregation requirements, and audit rights, to ensure that competitively sensitive information is not shared, directly or indirectly, with competitors through the platform. For energy companies, this review should encompass trading platforms, market intelligence services, and any AI-driven forecasting tools that ingest proprietary operational data.

Integrate Antitrust Review into AI and Trading Governance Frameworks

Companies should update their existing AI governance frameworks to include a mandatory antitrust review before any algorithmic pricing, procurement, trading, or bidding tool is deployed or materially updated. This review should assess whether the tool facilitates the exchange of competitively sensitive information among competitors, whether the tool’s outputs could function as a mechanism for price coordination, and whether the company retains meaningful independent decision-making authority over pricing and bidding. The antitrust review should be documented and should involve both legal and compliance personnel with antitrust expertise.

Expand Antitrust Training to Technical and Trading Personnel

Glad’s remarks signal that the DOJ views a broader range of employees as having potential personal criminal exposure in algorithmic cases, including not only executives and sales personnel, but also engineers, data scientists, product managers, and account managers. For energy companies, this training audience should include quantitative analysts, algorithmic traders, dispatch operators, and software developers working on bidding and pricing systems. Training should cover the specific risks of algorithmic collusion, including how shared data inputs and coordinated outputs can constitute an antitrust violation, even absent direct communication with competitors. Training should also address the Whistleblower Rewards Program and the company’s internal reporting mechanisms to ensure employees understand their obligations and options.

Establish Rapid Self-Reporting Protocols

The dual-race dynamic created by the Whistleblower Rewards Program and the Leniency Program demands that companies establish expedited internal protocols for evaluating potential antitrust violations. Glad’s warning that “a company’s path to Leniency can close while its general counsel is scheduling the next meeting” should be taken seriously. Energy companies, which often operate in fast-moving trading environments with large numbers of personnel who have access to competitive data, should pre-designate an internal team with authority to evaluate potential antitrust issues on a compressed timeline, establish clear escalation procedures, and maintain an up-to-date understanding of the Leniency Program’s requirements. Companies should also consider establishing formal whistleblower and internal reporting channels that incentivize employees to raise antitrust concerns internally.

Document Independent Decision-Making

Companies that use algorithmic pricing and trading tools should document that they retain independent authority over pricing decisions, including the ability to override algorithmic recommendations. Internal policies should require that pricing and trading teams record the basis for pricing decisions, particularly when those decisions follow algorithmic outputs. Companies should also ensure that their use of algorithmic tools does not create a pattern of “follow-the-algorithm” behavior in which pricing recommendations are adopted without independent business judgment. In the energy context, this documentation could be critical in defending against future enforcement actions or civil claims.

Monitor Regulatory and Judicial Developments

The legal landscape surrounding algorithmic pricing and AI-driven collusion is evolving rapidly. In-house and outside counsel should closely monitor ongoing judicial developments as well as any future enforcement actions, guidance, or policy updates from the Antitrust Division. Energy companies, as participants in concentrated industries with repeated competitor interactions, shared pricing platforms, and trade association participation, should conduct periodic antitrust risk assessments to ensure that their compliance programs keep pace with the DOJ’s enforcement posture.