AI Pricing and Competition Law: Algorithms, RealPage, and FTC Section 5

When does algorithmic pricing become an antitrust problem?

Algorithmic pricing becomes an antitrust problem when competitors use a shared tool to coordinate prices. A company may set its own prices any way it wants, including with software.

The legal hook is the agreement. Section 1 of the Sherman Act bars agreements that unreasonably restrain trade. Antitrust enforcers have argued that a shared pricing tool can carry such an agreement when competitors feed it nonpublic data and then follow its recommendations.

The first appellate decision in the recent wave of algorithmic pricing cases, as Baker McKenzie described it, went against plaintiffs. In Gibson v. Cendyn Group, decided August 15, 2025, the Ninth Circuit affirmed dismissal of claims that Las Vegas hotels violated Section 1 by separately licensing the same pricing software. The court held that alleging competitors independently bought licenses for a tool that recommends prices was not enough to state a claim.

The Justice Department had filed an amicus brief supporting the plaintiffs. The federal enforcers take a broader view of the theory than that court did, so a dismissal in one circuit does not settle the risk. That risk sits in the data that goes into the tool and in how much discretion users keep over the price that comes out.

What did the RealPage settlement require?

The RealPage settlement limits what data RealPage’s models may train on and how finely they may treat geography, and it includes no money damages.

The Justice Department sued RealPage and several large landlords in August 2024, alleging that RealPage’s rent-setting software let landlords share nonpublic data and align rents. On November 24, 2025, the Department filed a proposed final judgment with RealPage in the Middle District of North Carolina.

According to the Department and law firm summaries, the settlement:

• Limits the sharing of competitively sensitive, nonpublic information among competing landlords through RealPage
• Restricts the data used to train RealPage’s models to historical data at least 12 months old
• Bars models that identify geographic effects more narrowly than at a state level, and bars sharing other owners’ nonpublic data with users in any form, aggregated or not
• Requires removing or redesigning features that limited rent decreases or aligned pricing among users, including auto-accept settings that users do not set themselves
• Requires RealPage to cooperate in the case against landlords that used its software
• Includes no money damages and no admission of wrongdoing

Press reports in May 2026 said the court approved the settlement. The Department has since announced proposed settlements with landlord defendants, including Pinnacle on September 4, 2026. The state attorneys general who joined the federal case continue their claims, and on September 30, 2026 the court denied motions to dismiss filed by RealPage and several landlords. Private class actions continue separately.

For any business that uses a shared pricing vendor, the settlement terms read as a working checklist of what enforcers find acceptable: older data, no sharing of competitors’ nonpublic data with users, coarse geography in the models, and no feature that pushes users toward the recommended price.

What have states enacted?

California and New York have written common pricing algorithms into their antitrust laws. California makes it unlawful to use or distribute one as part of a contract, combination, or conspiracy in restraint of trade, or while coercing another person to adopt its recommended price or commercial term. New York bars landlords from knowingly or recklessly setting rents or lease terms based on recommendations from coordinating software, although a federal court has preliminarily enjoined that provision.

California AB 325

Signed October 6, 2025 and effective January 1, 2026, AB 325 adds Section 16729 to the Cartwright Act, California’s antitrust law. It makes it unlawful to:

• Use or distribute a common pricing algorithm as part of a contract, combination, or conspiracy in restraint of trade
• Use or distribute a common pricing algorithm while coercing another person to adopt its recommended price or commercial term

A “common pricing algorithm” is any methodology, including software, used by two or more persons that uses competitor data to recommend, align, stabilize, set, or otherwise influence a price or commercial term.

The law also lowers the pleading standard for Cartwright Act conspiracy claims generally, not only algorithm cases. Legislative committee notes describe this as a rejection of the stricter federal standard from Bell Atlantic v. Twombly. A companion bill, SB 763, raised Cartwright Act penalties.

Two points are still open. The statute does not define “coercion,” and commentators expect litigation over whether a discount for following the recommended price counts. And the lower pleading standard means a California complaint over a shared pricing tool is more likely to survive a motion to dismiss than the federal complaint in Gibson did.

New York rent algorithm law

New York enacted S.7882 (Chapter 437 of 2025), adding General Business Law Section 340-b to the Donnelly Act. Signed October 16, 2025 and effective December 15, 2025, it bars knowingly or recklessly facilitating an agreement among landlords not to compete, including by operating or licensing software that performs a coordinating function, and bars landlords from knowingly or recklessly setting rents or lease terms based on that software’s recommendations (Section 3). RealPage challenged the law. On September 29, 2026, the federal court in Manhattan granted RealPage a preliminary injunction against enforcement of Section 3 and denied the Attorney General’s motion to dismiss (RealPage, Inc. v. James, No. 1:25-cv-09847, S.D.N.Y.). A motion for reconsideration was filed October 13, 2026. Check the docket before relying on Section 3.

Several cities had passed similar bans earlier, including San Francisco (Ordinance 224-24, 2024) and Seattle (Council Bill 121000, June 2025). Press reports describe similar ordinances in Philadelphia and Jersey City.

What is surveillance pricing, and what rules apply?

Surveillance pricing is setting different prices for different people based on personal data about them, such as location, browsing history, or a profile.

The FTC opened a study of the practice in 2024. On January 17, 2025, it released initial staff research summaries finding that intermediaries used a wide range of personal data to target prices. The Commission voted 3 to 2 to release them, with two commissioners dissenting. No final report had been issued as of this writing.

New York acted by statute. The Algorithmic Pricing Disclosure Act, General Business Law Section 349-a, took effect November 10, 2025, after a federal judge dismissed the National Retail Federation’s First Amendment challenge on October 8, 2025.

• A business that sets a price with an algorithm using a consumer’s personal data must display, near the price: “THIS PRICE WAS SET BY AN ALGORITHM USING YOUR PERSONAL DATA.”
• Civil penalties of up to $1,000 per violation, enforced by the Attorney General after a cure process
• Exemptions include certain insurers, financial institutions subject to federal privacy law, and some subscription pricing

The Attorney General has since sent at least one company a letter about its pricing practices under the law.

How does FTC Section 5 apply to AI claims?

The FTC applies Section 5 of the FTC Act, which bars unfair or deceptive acts or practices, to AI claims directly, without an AI-specific rule.

• Rite Aid (December 19, 2023): The FTC alleged Rite Aid used facial recognition in stores without reasonable safeguards, and that it falsely flagged customers, particularly women and people of color, as shoplifters. The order bans Rite Aid from using facial recognition for surveillance for five years and requires deletion of images and of data, models, or algorithms derived from them.
• Operation AI Comply (September 25, 2024): Five actions against AI-related deception, including DoNotPay’s “robot lawyer” claims (final order in February 2025 with $193,000 in monetary relief) and several business-opportunity schemes promising AI-driven income.
• Rytr set aside (December 2025): The FTC reopened and vacated its 2024 consent order against Rytr, a writing tool accused of enabling fake reviews, saying the complaint did not adequately plead a violation and citing the administration’s AI Action Plan.
• Proposed policy statement (July 2026): Under Executive Order 14365, the FTC proposed a statement on how Section 5 applies to AI accuracy. It treats deliberately steering a model’s outputs toward undisclosed objectives as potentially deceptive, says ordinary hallucinations alone do not raise Section 5 concerns, and takes the position that a state law requiring an AI company to deceive consumers, for example by altering truthful outputs, is impliedly preempted to the extent it conflicts with the FTC Act. Comments closed July 31, 2026.

The pattern under current leadership is clear enough to plan around. False claims about what AI does still draw cases. Theories that a tool might be misused by others draw fewer.

What should a business using pricing software check?

Check where the vendor’s training data comes from, what competitor information its outputs reveal, and who makes the final price decision. Those answers separate lower-risk use from higher-risk use.

• Get the vendor’s written description of training data sources and freshness
• Turn off auto-accept features and track how often staff override the recommendation
• Do not discuss the tool’s recommendations with competitors
• Review vendor marketing for language about pricing discipline or avoiding price competition
• Add the New York disclosure wherever personalized algorithmic prices reach New York consumers
• Review every public claim about your own AI products for accuracy

What is settled, and what is still open?

The settled rule is old: competitors may not agree on prices, directly or through a tool. What is still being decided, in settlements, state statutes, and early appellate decisions, is how much data sharing and how much automation turns a tool into that agreement.

See the state law guide and the AI compliance guide for related duties.

This guide is general information as of October 2026, not legal advice. Antitrust questions turn on facts; consult antitrust counsel before changing how prices are set.