Algorithmic Pricing: New DAI Case, First Appellate Decision, and the Greystar Settlement. The show Goes On! (Part 1)

Author: Luis Blanquez

Summer is over and everyone is back at the office. If you’ve been enjoying some days off, you’ve probably missed what happened recently in the algorithmic-pricing space in the US. And, as always, we had a very busy summer here.

First, there is a new case from the Northern District of California that confirms the importance of not only the nature of the information shared with the software provider, but also whether such information is later incorporated into any final pricing recommendations.
Second, we have the first opinion from a Court of Appeals—in this case the Ninth Circuit in the Gibson v. Cendyn Group case—analyzing an antitrust claim involving algorithmic pricing.
And third, the DOJ Antitrust Division announced its recent settlement with Greystar Management Services to resolve the DOJ’s claims about the use by Greystar of RealPage’s software.

As I said, everyone has been working hard around here during the past months!

This first article explains the Dai v. SAS Inst. Inc new case. Also, if you need some background on the current cases before diving into the new developments, we’ve written several articles on algorithmic pricing:

Dai v. SAS Inst. Inc., No. 24-CV-02537-JSW (N.D. Cal. July 18, 2025)

In this new case, plaintiffs sued software provider IDeaS, Inc. (“IDeaS”), and a group of hotel operators including Wyndham Hotels & Resorts, Inc., Hilton Domestic Operating Company Inc., Four Seasons Hotels Limited, Omni Hotels Management Corporation, and Hyatt Corporation, for conspiring to fix hotel room prices (“Hotel Operators”).

Here are the allegations:

IDeaS is the dominant provider of revenue management and profit optimization software and services for Hotel Operators.

According to the complaint, Hotel Operators agreed to provide IDeaS with non-public, competitively sensitive price and occupancy information in real time, including the price paid by consumers for each room, the quantity of rooms available by room type, whether or not any consumers attempted to book a room that was no longer available, and room rates not visible to the public.

IDeaS would then plug all the information into its algorithm, generating supra-competitive pricing recommendations for each of them.

And the last—but certainly not least—piece of the puzzle: each defendant would implement IDeaS’s supracompetitive pricing, because they know all their horizontal competitors are doing the same thing.

Plaintiffs did not rely on direct evidence but rather alleged the inference of a horizontal agreement by the group of Hotel Operators using IDeaS’s software. They argued parallel conduct—when Hotel Operators began to use IDeaS’s software and to charge allegedly supra-competitive rates based on IDeaS’s recommendations—together with (i) an invitation to collude as well as the motive and the opportunity to do so; (ii) high barriers to enter the relevant market; (iii) inelastic demand for hotel rooms; and (iv) sharing of confidential information against self-interest; all as plus factors to show antitrust conspiracy.

Would this be enough to show the existence of an antitrust conspiracy? Not quite, according to the Northern District of California.

Remember, Sherman Act Section 1 antitrust cases require: (1) a contract, combination, or conspiracy among at least two different entities; (2) an intent to restrain trade; and (3) injury to competition. See here and here.

On July 18, 2025, the District Court for the Northern District of California dismissed the complaint on several grounds. What is particularly helpful for future litigants in this Opinion is the comparison the Court makes with past recent cases.

The Court in California held here that plaintiffs did not sufficiently allege parallel conduct for several reasons.

First, the Opinion compares the current case to In re RealPage, Inc., Rental Software Antitrust Litig., 709 F. Supp. 3d 478 (M.D. Tenn. 2023), where a critical level of RealPage’s software adoption explaining how defendants changed their strategy and increased prices—despite not acting simultaneously—was enough to show parallel conduct.

Then, and in contrast, it mentions Gibson v. MGM Resorts Int’l, No. 2:23-cv 00140-MMD-DJA, (D. Nev. Oct. 24, 2023), where a court dismissed the complaint for lack of parallel conduct. In that case, plaintiffs neither include information about when the defendants began to use the software and which systems they used, nor alleged facts about the rate at which the defendants accepted the software recommendations. Plaintiffs did include general allegations of the acceptance rate for the price recommendations, but that was not sufficient to make the existence of an agreement plausible according to Twombly requirements.

Last, the Northern District of California states that plaintiffs did not provide enough facts in this case to explain when the Hotel Operators began to outsource their pricing decisions to IDeaS; when they started to change their strategy and increased prices; or when and how they started to adopt IDeaS’s pricing recommendations to their room prices.

In other words, according to the Court, plaintiffs did not have to show that each defendant acted at the exact same moment in time or acceptance rate. But plaintiffs did have to plead additional facts to render the allegations of parallel conduct plausible, which as explained below, they didn’t do.

Indeed, the Court reasoned that plaintiffs did not allege enough plus factors:

First because plaintiffs did not include any facts to explain how the COVID-19 pandemic impacted them—rather than the hotel industry in general—or why such a motive would be more indicative of collusion than market interdependence.

Second, plaintiffs also alleged as a plus factor that IDeaS hosted events and summits for its users. But again, the Court concluded that plaintiffs did not include any facts showing any of the hotel defendants attending one of those events. Or even if they did, any facts explaining what agreements were made, if any.

Last, and this is the most interesting discussion in the Opinion, plaintiffs argue that the Hotel Operators acted against their self-interest by providing IDeaS with their confidential information, knowing it would be shared with their competitors. Indeed, the exchange of confidential information has been a decisive plus factor in past decisions involving algorithmic pricing. And the Northern District of California takes this opportunity to provide a very helpful analysis from past cases:

First the Court distinguishes this case from the two cases surviving a motion to dismiss in the past. In Real Page, the court reasoned that the most persuasive evidence of horizontal agreement was the fact that each defendant provided RealPage its proprietary commercial data, knowing that RealPage would require the same from its horizontal competitors and use all of that data to recommend rental prices to its competitors. Same as in Duffy, where plaintiffs alleged the software worked only if each defendant divulged confidential and commercially sensitive pricing information and de-prioritized occupancy in favor of algorithmic pricing.
Then, the Court discussed two cases that were dismissed because plaintiffs failed to plausibly allege exchange of confidential information. First, In Cornish-Adebyi, the Casino-Hotels defendants knowingly provided their non-public room and pricing and occupancy data to the software. But plaintiffs’ allegations failed to show that the data was pooled or comingled into a common dataset against which the algorithm was running. In Gibson v. MGM Resorts, the court reasoned that the plaintiffs’ allegations showed “confidential information is fed in, but less clearly out, of the algorithms”, finding allegations insufficient to plausibly allege use and exchange of confidential information.

After this analysis and looking at each of the plus factors individually and holistically, the Court in DAI concluded that this case was more analogous to Cornish-Adeybi and the Gibson cases for the following reasons: (1) there were no facts to make it reasonable for the Court to infer that the pricing information was based on non-public sources; and (2) plaintiffs were able to successfully show how IDeaS’s software plugged the Hotel Operators’ information into the algorithm, but not how such information was later included in any pricing recommendations.

This new case confirms the high standard required to survive a motion to dismiss with antitrust claims involving algorithmic pricing. Not only is it important to provide enough facts about the nature of the information shared with the software provider and the way defendants share such information, but also whether that same information was later commingled and integrated into final pricing recommendations. This is consistent with what the Ninth Circuit has recently held in the first appellate decision in the US dealing with algorithmic pricing, and the Greystar Settlement. We analyze both in the next article.

Image by Leo from Pixabay