S. 232 Senate Bill 119th Congress

Preventing Algorithmic Collusion Act (S.232) — distribution to two or more in the same market presumes agreement under the Sherman Act, rebuttable only by clear and convincing evidence

U.S. Senate Latest update Jan 23, 2025

A Senate bill requiring reports on pricing algorithms on written request from the Attorney General or the Commission, and presuming agreement under section 1 of the Sherman Act in defined circumstances. Rebutting the presumption requires showing both that the defendant neither developed nor distributed the algorithm and that it did not and could not reasonably have known nonpublic competitor data was used, by clear and convincing evidence.

Bill overview (primary data)

  • Bill numberS. 232
  • TypeSenate Bill
  • Congress119th Congress
  • Latest actionRead twice and referred to the Committee on the Judiciary.(2025-01-23)

Key points

  • Requires a written report on a pricing algorithm within 30 days of a request by the Attorney General or the Commission.
  • The report covers autonomy and human review, the rules or processes used, the data and its sources and frequency, and price discrimination.
  • Presumes agreement under section 1 of the Sherman Act where an algorithm was distributed to or used by two or more in the same or a related market.
  • Rebuttal requires neither developing nor distributing it and showing, by clear and convincing evidence, no actual or constructive knowledge of nonpublic competitor data.
  • Joint and several liability applies in civil cases where the presumption operates.
  • Agreement is presumed from distribution to two or more in the same or a related market, rebuttable only by clear and convincing evidence.

1Moving the burden of proof

Proving collusion is held to be hard because the existence of an agreement must be shown. Where an algorithm mediates, evidence of people talking often does not exist. This bill moves that burden by placing a presumption that an agreement existed.

The presumption operates where an algorithm was distributed to two or more persons with the intent that it set or recommend prices in the same or a related market, among other cases. The starting point is not that anyone talked but that the same tool was handed to competitors.

2The conditions for rebuttal are set high

Overturning the presumption requires two things: that the defendant neither developed nor distributed the algorithm, and that it did not have actual knowledge and could not reasonably have known that nonpublic competitor data was used. The second must be shown to the high standard of clear and convincing evidence.

For the party that built the tool or handed it out, this route is not open at all. For a user, saying it did not know the contents is not enough. Not only is a presumption placed; how it may be overturned is finely designed.

3The reporting duty opens the contents

Section 3 requires a report within 30 days of a written request by the Attorney General or the Commission. What must be included is concrete: whether prices are set autonomously or with human review, what rules or processes produce a price, which data are used and from where and how often they are collected, and whether price discrimination occurs.

The design compels an explanation of a mechanism invisible from outside. Of the 114 AI-related bills this site holds as of 2026-09-01, sorting the 44 not yet covered by the leading verb of the official title gives establish at 11, require at 9, direct at 6 and amend at 6, while this is the only one beginning with prevent.

4Moving the burden of proof

Proving collusion is difficult because agreement has to be shown. Where an algorithm mediates, evidence of people conferring often does not exist. This bill places a presumption there.

  1. 1Proving collusion ordinarilyAgreement itself has to be shown
  2. 2Where an algorithm mediatesEvidence of people conferring often does not exist
  3. 3Place a presumptionTake as the trigger that the algorithm was distributed to two or more in the same or a related market
  4. 4Narrow the rebuttalShow, by clear and convincing evidence, no development or distribution and no actual or constructive knowledge that non-public competitor data was used

Those who built or distributed the tool have no route to that rebuttal at all. Those who used it cannot simply say they did not know what was inside. Section 3 requires a report within 30 days of a request by the Attorney General or the Commission, covering whether prices are set autonomously or with human review, and what data is collected, from where and how often.

Why it matters

The same pricing algorithm handed to competitors can produce aligned prices without anyone talking. This bill takes that state itself as the starting point for presuming agreement and narrows the path to rebuttal. For a provider of such tools, records of who received it and what data it used would carry legal weight.

FAQ

Why place a presumption?
Proving collusion requires showing an agreement, and where an algorithm mediates, evidence of people talking often does not exist.
How can the presumption be rebutted?
By neither developing nor distributing the algorithm and showing, by clear and convincing evidence, no actual or constructive knowledge that nonpublic competitor data was used.
What goes in the report?
Responsibility for development or distribution, autonomy and human review, rules and processes, data and collection, and whether price discrimination occurs.

Sources (primary)

Source: Congress.gov (Library of Congress; U.S. legislative materials, public domain). Links go to the official site.

#Congress#Bills#Algorithms#Antitrust#Pricing
Disclaimer: This site independently summarizes and classifies information based on official data sources. Always verify the latest and accurate information with the official sources. Content on finance, health, legal, and security is information, not advice. This site is not an official website of the U.S. government.