Preventing Algorithmic Collusion Act of 2025
The Preventing Algorithmic Collusion Act of 2025 targets the use of pricing algorithms (including those driven by machine learning or AI) that rely on nonpublic competitor data to set or recommend prices or commercial terms. The bill prohibits such algorithms, establishes a framework for antitrust enforcement including an audit/reporting mechanism for algorithm use, introduces a transparency requirement for large price-setting firms, and creates a legal presumption that certain algorithmic pricing constitutes unlawful price fixing under the Sherman Act and FTC Act. It also directs a federal study on pricing algorithms to inform future policy. In short, the bill aims to curb algorithm-driven collusion and to increase oversight and accountability for pricing technologies. Key features include mandatory audits and reporting to the Attorney General/FTC, civil penalties for using pricing algorithms that rely on nonpublic competitor data, a legal presumption of collusion in certain algorithmic pricing scenarios (with a narrow rebuttal), mandatory disclosures for large firms, and a forthcoming FTC study on prevalence, effects, and regulatory needs.
Key Points
- 1Prohibition and penalties for pricing algorithms that use nonpublic competitor data: It is unlawful to use or distribute such algorithms, with civil penalties and potential injunctive relief for violations. Effective 90 days after enactment.
- 2Competition law enforcement audit tool: Requires entities using or distributing pricing algorithms to provide a written, detailed report to the Attorney General or the FTC within 30 days of a request. Reports cover development/distribution, autonomy of pricing decisions, data inputs (including training data), data collection processes, price discrimination, and changes to the algorithm. Reports are confidential.
- 3Presumption of unlawful price fixing: Creates a legal presumption that certain uses of pricing algorithms constitute an agreement or conspiracy in restraint of trade under the Sherman Act and an unfair method of competition under the FTC Act, when the algorithm is distributed or used to set prices in the same market and nonpublic competitor data is involved. There is a narrow rebuttal path if the defendant can show no knowledge of the use of nonpublic competitor data.
- 4Transparency for large users of pricing algorithms: Firms with $5 million+ in annual revenue that use pricing algorithms must disclose to customers and to employees/contractors when prices are algorithmically set or recommended. Disclosures may include information about price discrimination and whether a third party developed the algorithm. Noncompliance can trigger FTC enforcement.
- 5FTC study: The Commission must publish a study within 2 years on prevalence, wage/price discrimination, potential harms and benefits of pricing algorithms, and recommendations for further legislation or rulemaking.