Preventing Algorithmic Collusion Act of 2025
Sponsored By: Senator Amy Klobuchar
Introduced
Summary
Ban on pricing tools trained with rivals' secret data. This bill would bar using or distributing pricing algorithms that are trained with nonpublic competitor data and would create new transparency and enforcement rules to deter algorithmic collusion.
Show full summary
- Consumers and households would get advance notice before purchase when a pricing algorithm set or recommended a price and a warning that the algorithm may treat customers differently; companies with at least $5 million in annual revenue face penalties for failing to disclose.
- Workers and contractors would be told when prices or terms that affect them are set by an algorithm and warned if the algorithm may discriminate across employees or contractors.
- Businesses that use or distribute pricing algorithms would have to produce a certified, detailed report to the Department of Justice or the Federal Trade Commission on request, typically within 30 days, describing rules, training data, developers, and whether the tool acts autonomously. Using or distributing an algorithm trained with nonpublic competitor data would be unlawful and subject to civil penalties of at least $10,000 per day or other equitable relief, and a presumption of collusion can arise when the same algorithm is used by multiple firms; reported material would be kept confidential and may be shared with NIST for technical assistance.
Personalized for You
How does this bill affect your finances?
Personalize government policy and PRIA will tell you what this bill means for your household, plus every other piece of legislation we track. PRIA reads each provision against your financial profile to show you exactly what matters to your wallet.
Bill Overview
Analyzed Economic Effects
4 provisions identified: 1 benefits, 0 costs, 3 mixed.
Required algorithm notices to users
If enacted, businesses with $5,000,000 or more in yearly revenue that use a pricing algorithm would have to tell customers before purchase that prices or terms are set or recommended by an algorithm. They would also have to tell current or prospective employees or contractors when pay or terms are set or recommended by an algorithm. If the algorithm treats different customers or workers differently, firms must say so and name a third-party developer if one supplied the algorithm. Failure to disclose could trigger FTC penalties of at least $5,000 per day (inflation adjusted) and other relief.
Ban on competitor-data pricing algorithms
If enacted, the bill would make it unlawful to use or distribute any pricing algorithm that was trained on nonpublic competitor data. That ban and a related legal presumption would take effect 90 days after enactment. The FTC or Attorney General could sue and seek either at least $10,000 per day (inflation adjusted) or the price of products sold using the violating algorithm, plus injunctions. The bill would also create a rebuttable presumption of collusion if an algorithm is given to or used by multiple firms in the same market; distributors who knew or should have known about the nonpublic data could face joint and several liability.
Mandatory algorithm audit reports
If enacted, the Attorney General or the FTC could require any person who uses or distributes a pricing algorithm to give a written, certified report. The report would be due within 30 days (or later if the agency allows). It would have to name developers and responsible officers, say whether the algorithm acts autonomously, describe the rules and all training data and sources, and disclose any price discrimination. A senior officer must sign under penalty of perjury. The reports would be treated as trade secrets and could be shared between DOJ, FTC, and NIST for technical help.
What counts as a pricing algorithm
This bill would define key terms about pricing algorithms. It would say pricing algorithms include machine learning or AI that set or recommend prices or other commercial terms in interstate commerce. It would define "nonpublic data" and "nonpublic competitor data" and list narrow exclusions. The bill would also say "price" can include money paid to employees or contractors.
Sponsors & CoSponsors
Sponsor
Amy Klobuchar
MN • D
Cosponsors
Sen. Wyden, Ron [D-OR]
OR • D
Sponsored 1/23/2025
Sen. Durbin, Richard J. [D-IL]
IL • D
Sponsored 1/23/2025
Peter Welch
VT • D
Sponsored 1/23/2025
Sen. Hirono, Mazie K. [D-HI]
HI • D
Sponsored 1/23/2025
Sen. Luján, Ben Ray [D-NM]
NM • D
Sponsored 1/23/2025
Sen. Shaheen, Jeanne [D-NH]
NH • D
Sponsored 1/23/2025
Sen. Murphy, Christopher [D-CT]
CT • D
Sponsored 1/23/2025
Sen. Blumenthal, Richard [D-CT]
CT • D
Sponsored 1/23/2025
Roll Call Votes
No roll call votes available for this bill.
View on Congress.gov