I. Introduction and Summary

On February 23, 2026, the Department of Justice’s Antitrust Division (DOJ) and the Federal Trade Commission (FTC) launched a joint public inquiry regarding potential guidance on collaborations among competitors. The request for information will help the agencies to develop new guidance after the 2000 Antitrust Guidelines for Collaborations Among Competitors[1] (Collaboration Guidelines) was withdrawn by the agencies in December 2024.

In the announcement of the withdrawal, the FTC and DOJ advised “businesses considering collaborating with competitors…to review the relevant statutes and caselaw to assess whether a collaboration would violate the law.”[2] With many firms lacking antitrust expertise to simply “review the relevant statutes and caselaw,” the withdrawal of the Collaboration Guidelines created uncertainty for businesses seeking to collaborate and trade associations that often act as an intermediary in these procompetitive ventures.

Collaborations are often procompetitive. They enable firms to reach new markets, develop new products, or engage in research and development that one firm alone cannot achieve. Without proper guidance, firms may be hesitant to engage in such projects, robbing consumers of potential benefits.

While these comments are not exhaustive, they focus on several areas of current concern: algorithmic pricing, information sharing, and artificial intelligence (AI) research and development collaborations.

II. Algorithmic pricing

Updating guidance on collaborations among competitors presents an opportunity to reflect the growing adoption of algorithmic pricing tools. These tools have raised antitrust concerns in instances where groups of competitors are using a common pricing algorithm.[3]

Typical antitrust principles should apply to such tools. When these algorithmic pricing tools are used by competitors to artificially inflate prices or restrict output in violation of Section 1 of the Sherman Act, the agencies should act.

Collaboration guidelines should clarify what sort of practices raise competitive concerns. In the case Duffy v. Yardi Systems, Inc., the plaintiffs claimed that competing landlords used Yardi’s pricing algorithm to “artificially inflate” multifamily rental prices.[4] In other words, the competitors allegedly engaged in a price fixing scheme. The DOJ and FTC published a joint statement of interest in the case outlining the legal principles that should be applied to claims of algorithmic price fixing. Referencing an earlier statement of interest filed by the DOJ in Re: RealPage, Rental Software Antitrust Litigation, the DOJ explained that it is per se unlawful “for competitors to join together their independent decision-making power to raise, fix, peg, or stabilize prices” under Section 1 of the Sherman Act. The DOJ also claimed that landlords were sharing real-time, non-public, and competitively sensitive data including “actual rents paid, occupancy rates, and records of lease transactions” to train the RealPage pricing model.

The agencies can use this example to create guidance by clarifying:

i. The type of information that can be used to train a common algorithm pricing model: Training a model on publicly available information compared to competitively sensitive information will foster a more efficient market while limiting firms’ ability to collude.

ii. Degree of autonomy over decision making: The price produced by the algorithm does not require mandatory adoption but is rather a recommendation that leaves an individual firm to make its own decisions.

iii. The time lag of the input data: Real-time data sharing could better facilitate collusion among users compared to data inputs that are reported with a time lag.

The agencies can also clarify that algorithmic pricing tools built using proprietary company data – with or without other publicly available data – for the sole use of that company is unlikely to raise competitive concerns. Applying algorithmic pricing tools can generate efficiencies for firms. A recent article noted that “30% of food in American grocery stores is thrown away each year, and some experts estimate that translates to nearly $18.2 billion in lost value.” Flashfood is a company “helping grocers reduce prices on items as they near the end of their shelf life.” Partners of Flashfood include Kroger, Piggly Wiggly, Loblaws, and Gelson’s, which have reduced these losses by 27 percent.[5]

III. Information Exchanges

Information exchanges among competitors raise concerns like those of common algorithmic pricing tools. Information exchanges – which are often run by third-party consultants and trade associations – conduct industry surveys and provide the aggregated market data to their customers. These surveys often include data on production, costs, salaries, benefits, revenue, and quantities sold. The data are used to help firms compete more effectively.

The recent settlement between the DOJ and Agri Stats, however, highlights the need for guidance with respect to information sharing among competitors.[6] Agri Stats, a meat processing industry information exchange, allegedly disseminated competitively sensitive information that enabled meat producers to raise prices and reduce output in violation of Section 1 of the Sherman Act.[7]

Similar to the withdrawal of the Collaboration Guidelines, the DOJ – followed by the FTC – withdrew the Antitrust Enforcement Policy Statements Issued for Health Care Industry (1993).[8] The 1993 policy statement established information exchange antitrust safety zones specific to the health care industry, yet the reach of the document extended to other industry information exchanges. The safety zones were expanded in the Statements of Antitrust Enforcement Policy in Health Care (1996) policy statement, from which the DOJ and FTC also withdrew.

While the allegations against Agri Stats show that the firm likely operated outside the antitrust safety zones outlined in the 1996 policy statement, firms are currently left with no guidance at all. The agencies can clarify:

i. The use of a third party: Most information exchanges operate using a third-party intermediary to avoid the direct sharing of information among competitors. Moreover, clarification on how a third-party intermediary can use both the aggregated data it ultimately shares with information exchange participants and the disaggregated data submitted by individual companies for their own commercial purpose.

ii. The type of information shared: Firms submitting historical information with a sufficient lag should raise little competitive concern. Establishing a recommended time lag based on the frequency of the underlying data should alleviate competitive concerns. Data sharing with respect to future production and prices, however, could afford firms the opportunity to collude. The sharing of non-economic data is unlikely to raise competitive concerns.

iii. Anonymization techniques: It was alleged that Agri Stats “loosely anonymized” its reports, simplifying a firm’s ability to unmask the identity of a competitor. Providing guidance that inhibits a firm’s ability to discern the information of another competitor is paramount to protecting competition. For example, creating a rule establishing the maximum market share one firm can have, a required minimum number of firms submitting data, or a combination of both for a particular line item of an information exchange report for it to be released to competitors can help avoid anticompetitive practices. Such clarification can be classified as an “antitrust safety zone.”

Clear guidance will help competitors avoid raising antitrust concerns while also protecting third-party intermediaries from mistakenly facilitating a collusive environment.

IV. Artificial intelligence research and development collaborations

The emergence and rapid expansion of AI has led to headlines marking joint ventures within the same or across levels of the technology stack. Multi-billion-dollar partnerships between Microsoft and OpenAI, Google and Anthropic, and others suggest competition is intense. These coalitions have proved fluid, however, as structural changes to agreements continue to be made as the industry continues to evolve and the needs [9] each firm changes.

AI is a nascent industry operating with few regulatory guardrails. Keeping these barriers to a minimum helps foster the industry’s rapid innovation as AI leadership is in constant flux, often with a new dominant player being replaced as quickly as it was crowned.

The past three years have featured a tremendous buildout in AI infrastructure including grid and energy procurement, data centers, and chips. Collectively, Amazon, Alphabet, Microsoft, and Meta are projected to spend $700 billion on infrastructure in 2026.[10] Yet many of these projects are being done by one company for its own use. Collaboration among competitors could make these transactions more cost efficient than one company building alone, and once completed, allow companies to produce more output, likely lowering prices. Guidance on such collaborations should outline instances where such joint ventures could run afoul of antitrust. Collaborative agreements conditional on activities that allocate markets, coordinate prices, or bid rigging are per se illegal, and absent evidence of such conditions, infrastructure collaborations should largely be considered procompetitive.

As the industry matures, it is likely that additional collaboration will be necessary to generate industry-wide standards governing data safety, model testing, and portability. Third-party intermediaries, including standards development organizations, will likely be involved in their creation and adoption and should be governed using the rule of reason standard.[11] Voluntary standards that promote interoperability, developed with transparency, and not designed to exclude firms or suppress innovation, are typically procompetitive. Standards often lower switching costs and provide consumers with important safety, operational, and technical information concerning products and services.

V. Conclusion

The FTC and DOJ should replace the withdrawn 2000 Antitrust Guidelines for Collaborations Among Competitors with a new document that reflects recent agency investigations and other areas of concern. Moreover, these updated guidelines would provide smaller firms that lack regulatory expertise to avoid running afoul of antitrust law.

[1] https://www.ftc.gov/sites/default/files/documents/public_events/joint-venture-hearings-antitrust-guidelines-collaboration-among-competitors/ftcdojguidelines-2.pdf

[2] https://www.ftc.gov/news-events/news/press-releases/2024/12/ftc-doj-withdraw-guidelines-collaboration-among-competitors

[3] https://www.americanactionforum.org/insight/primer-pricing-algorithms-and-antitrust/

[4] https://www.ftc.gov/system/files/ftc_gov/pdf/YardiSOI-filed%28withattachments%29_0.pdf

[5] https://www.cnbc.com/2026/04/17/some-grocers-are-using-ai-to-cut-food-waste-and-boost-profit-margins.html

[6] https://www.justice.gov/opa/pr/justice-department-requires-agri-stats-end-exchange-competitively-sensitive-information

[7] https://www.americanactionforum.org/insight/doj-revs-up-scrutiny-of-information-exchanges/

[8] https://www.justice.gov/archive/atr/public/press_releases/1993/211661.htm

[9] https://thenewstack.io/openai-microsoft-partnership-restructured/

[10] https://www.cnbc.com/2026/02/06/google-microsoft-meta-amazon-ai-cash.html

[11] https://www.ftc.gov/legal-library/browse/statutes/standards-development-organization-act-2004