DEFAULTING TO THE STATUS QUO: THE GOOGLE SEARCH REMEDIES DECISION

By Samuel Miller and Ryan Sandrock¹

The hardest part of a government-initiated Section 2 monopolization case is figuring out the appropriate remedy. In part, this is because the general pronouncements of the Supreme Court regarding remedies in monopolization cases are somewhat in conflict. On the one hand, antitrust relief should unfetter a market from anticompetitive conduct, "pry open a market that had been closed" by defendant’s illegal conduct and deny to the defendant the fruits of its statutory violation.² On the other hand, the remedy should not be designed to punish the defendant, should be "tailored to fit the wrong" and should not undermine the incentives of the defendant or rivals to innovate.³ Finding the right balance is especially difficult in cases, such as Microsoft and Google, where the defendant acquired monopoly power through innovation and "competition on the merits," but was found to have illegally maintained a monopoly by "exclusionary" means. And it is particularly difficult when the markets at issue change over the course of the case—so that a case about Search became a case about AI.

In the landmark decision of United States v. Google LLC, 747 F. Supp. 3d 1 (D.D. C. 2024), District Judge Amit Mehta found that Google had violated Section 2 of the Sherman Act by entering into actual or de facto exclusive dealing arrangements for default placement of Google search with browser developers such as Apple and Mozilla, wireless carriers such as Verizon and AT&T, and cellphone manufacturers such as Samsung and Motorola. The "wrong" committed by Google was to "lock up" the key channels of distribution for general search services to the exclusion of rivals such as Bing and DuckDuckGo, giving Google a massive scale advantage.⁴

This article will examine Judge Mehta’s recent Remedy Opinion, issued on Sept. 2, 2025, United States v Google LLC, case 1: 20-cv-03010, Dkt. 1436 (hereinafter Google Remedy Opn.) and discuss whether Judge Mehta found the right balance. Judge Mehta’s Liability and Remedy Opinions were heavily influenced by the decisions in the Microsoft case, including the historic D.C. Circuit en banc decision upholding Section 2 liability (but reversing the District Court’s remedy determination that Microsoft should be broken up), United States v Microsoft Corp., 253 F. 3d 34 (D.C. Cir. 2001), as well as the subsequent decision approving the consent judgment agreed to by the Antitrust Division and some

[Page 1]

states with Microsoft, New York v. Microsoft Corp., 224 F. Supp. 2d 76 (D.D.C. 2002) (hereinafter "New York") and the decision rejecting more severe remedies urged by other states, including California and Massachusetts, Massachusetts v. Microsoft Corp., 373 F. 3d 1119 (D. C. Cir. 2004) (hereinafter "Massachusetts").

Although Judge Mehta recognized that the general search market had been "frozen" for over 10 years, he noted that the recent emergence of generative AI ("GenAI) and chatbots had created a new and potentially serious competitive threat to Google’s dominance. Judge Mehta found that the GenAI space is "highly competitive", with new entrants and access to billions of dollars of capital.⁵ This clearly influenced his thinking as to which proposed remedies to accept and which to reject.

Judge Mehta appears to have faith that there will be an AI revolution that will disrupt current technologies and allow market forces to determine the next generation ofwinners and losers. Already, however, there are loud skeptics. As described in more detail below, the most jarring element of the opinion for many is that Google will still be allowed to pay for defaults. The liability decision concluded that those payments harmed competition and gave Google an unfair advantage, but the remedies decision seems to conclude that barring those payments could harm competition even more. Critics also are upset about the lack of any structural remedies, although a break-up was always more unlikely than a ban on defaults.

NO EXCLUSIVE DEALING AND LIMITING LICENSES TO A ONE YEAR TERM

As a starting point, the District Court ordered that Google should be prohibited from entering into exclusive licensing arrangements with browser developers, cellphone manufacturers, or wireless carriers, or agreements which conditioned the licensing of Google Play Store or other Google software on that entity’s distributing or preloading Google Search or Chrome.⁶ Interestingly, the District Court also limited the term of Google licensing agreements to one year. This limitation harkens back to the original Antitrust Division case against Microsoft in the mid-1990’s, in which the Consent Decree limited the term of Microsoft’s operating system licensing agreements to one year, in order to give rival operating system developers an opportunity to seek business from PC manufacturers⁷

However, Judge Mehta rejected Google’s argument that prohibiting exclusive arrangements is all that the Court could or should do, stating:

Merely excising the exclusive provisions from Google’s distribution agreements will not unleash competition. Google’s remedies fail to address any illegally obtained fruit of those agreements other that the

[Page 2]

‘freedom’ from competition it enjoyed for more than a decade. They do nothing to ‘eliminate’ the consequences of its exclusionary acts.⁸

In so ruling, Judge Mehta relied on the recent Ninth Circuit decision in Epic Games, Inc. v Google LLC (In re Google Store Antitrust Litig.), 2025 U.S. App. LEXIS 191185 (9^th Cir. July 31,2025), where the Court approved of remedies which go beyond a mere prohibition of anticompetitive conduct, and ordered Google to affirmatively grant third-party access to Google Play Store’s catalogue of apps and to allow such stores or platforms to be distributed through the Play Store.⁹

COURT REJECTS STRUCTURAL DIVESTITURE RELIEF AND CHOICE SCREENS

Significantly, Judge Mehta refused to accept Plaintiffs’ proposed requirement that Google divest the Chrome browser and related technologies. The Court noted that divestiture is a ‘drastic" remedy that should only be imposed with great caution. Further, Plaintiffs did not satisfy stricter causation standard for structural remedies, especially because the record contains ample evidence that lawful conduct ( i.e., that Google Search was "best-in-class") played an important role in the maintenance of Google’s monopoly.¹⁰For similar reasons, Judge Mehta rejected the proposal that the Final Judgment contain a provision allowing for the contingent divestiture of the Android operating system. The Court found that that this remedy "does not fit the wrong."¹¹

The Court also rejected the requirement that Google include a "choice screen" in its devices and browsers. A "choice screen" is a user interface that asks consumers to make an explicit choice among a variety of products. Although Judge Mehta recognized that a remedy which promotes consumer choice is consistent with the goals of antitrust law, he declined to impose such a remedy since he regarded it as a form compelled product re-design, which he found was not a proper task of the Court, citing New York, 224 F. Supp. 2d at 158.¹² Judge Mehta noted that the use of "choice screens" in Europe had not changed the competitive landscape.

COURT REJECTS PAYMENT BAN TO DISTRIBUTORS

One of the most controversial and hotly-contested remedy proposals was to prohibit Google from making search-related payments to distributors of Google Search, Chrome or Google’s new GenAI products, including payments for default or preferential placement as well as revenue share payments. The recipients of such payments, including Apple and Mozilla, objected to this proposal, since it would deprive them of significant revenue.¹³ Proponents of this remedy argued that it would create an opportunity for other search

[Page 3]

engine rivals to pay for and secure default placement, thus stimulating competition in the search engine market.

However, Judge Mehta rejected this remedy, finding that banning Google payments would pose a substantial risk of financial harm to browser developers, wireless carriers, and cellphone manufacturers, most of whom would still select Google search since it is best search provider¹⁴ Banning such payments would save Google money but could destroy smaller browser developers such as Mozilla and result in higher mobile phone prices and less innovative phone features. Balancing the competitive benefit against the potential consumer harm, Judge Mehta came down on the side of allowing Google to continue to make payments to distributors for default placement. As discussed, Judge Mehta was influenced by the recent emergence of GenAI as a threat to the primacy of traditional internet search.¹⁵

Some commentators have questioned whether allowing Google to continue to pay for default placement will undercut the effectiveness of other remedies and will allow Google to continue to dominate the search engine market. And some have questioned whether the remedy decision can be reconciled with the liability decision. Professor Erik Hovenkamp noted that Judge Mehta gave three arguments for not barring Google from paying for defaults—two of which were arguments Judge Mehta had rejected in the liability phase. Specifically, Judge Mehta, in the remedies opinion, accepted the arguments that: (1) Google’s payments would promote browser competition and increase smartphone competition and (2) allowing smaller competitors to obtain default distribution (instead of Google) would be harmful, because currently available search engines from smaller rivals are not as good as Google Search. But, in the liability opinion, Judge Mehta had rejected those arguments. So, if the default payments actually promoted competition, why did the Court find them to be illegal in the first place?¹⁶

DATA-SHARING REMEDIES

Plaintiffs proposed a variety of data-sharing requirements, including certain Search Index data, three sets of User-side data, and certain Ads data. The aim of these requirements was to provide rivals with more query data and analysis to overcome the massive scale advantage which Google had obtained as a result of its exclusionary distribution agreements. Judge Mehta agreed that some compelled data sharing was a reasonable method of eliminating the consequences of Google’s illegal action and well within the range of ordinary Sherman Act remedies.¹⁷ It was also consistent with the rulings in the Google Play Store case, in the which the Ninth Circuit affirmed a remedy requiring Google to share its catalogue of apps with rivals to overcome Google’s illegally amplified network effects by giving rivals a fair opportunity to establish themselves, and in the Microsoft case, in which the D.C. Circuit affirmed the compelled disclosure of APIs

[Page 4]

and communication protocols to enable interoperability between Microsoft’s Windows operating system and rival middleware vendors.¹⁸

However, after a lengthy analysis of the data-sharing remedy proposal, Judge Mehta narrowed the datasets to be disclosed to tailor the remedy to the nature of Google’s anticompetitive conduct¹⁹, and also limited the frequency of disclosure. The Court also declined to require disclosure of certain data-analytics tools which Google had developed for internal use, and certain data used to train Google’s GenAI tool. To protect Google’s incentive to innovate to avoid "cloning" of Google technologies, the data-sharing remedy involves no compelled disclosure of IP or trade secrets, such as algorithms, ranking signals, or post-trained LLMs used to deliver GenAI results²⁰ The Court also rejected the proposal to require disclosure of Ads data, finding that there was no showing that disclosing such data would increase competition in the general search text ads market.

SEARCH SYNDICATION REMEDIES

"Search Syndication" refers to a commercial arrangement whereby one search engine provider supplies to search services to another entity (e.g., Yahoo contracting with Bing to provide search services for Yahoo users.)

Plaintiffs had proposed that Google offer broad search syndication services to "qualified competitors" for 10 years at no more than "marginal cost". Judge Mehta agreed that some syndication remedy was appropriate, but Plaintiffs’ proposal must be narrowed. The rationale for some syndication remedy is that rival search engines will need some years to build up their own search indexes and algorithms to try to match Google’s quality, so such rival should be able to "borrow" Google’s better results for a while. But Judge Mehta ruled that the scope of syndication for both search engines and search text ads should be restricted, limited to a term of five years, and offered to rivals on ordinary commercial terms, not a "marginal cost"²¹

OTHER BEHAVIORAL AND ADMINISTRATIVE REMEDIES

The Court agreed with Plaintiffs that a Technical Committee should be established to aid the Plaintiffs and the Court in administering and enforcing the Final Judgment. The Technical Committee would advise regarding which rivals would be "Qualified Competitors"; recommend reasonable data security standards for such "Qualified Competitors"; advise on an appropriate cap on User-side Data disclosure; help formulate a tapering rate for search syndication; and other matters.

However, Judge Mehta rejected Plaintiffs” proposal that Google fund a "public education" campaign regarding the case and be required to give advance notice of any investment in a GenAI or search company, beyond what is already required under the Hart-Scott-Rodino Act ( which requires notification of mergers or acquisitions above

[Page 5]

$126.4 million). The Court also rejected proposed language characterized as "anti-circumvention" and "anti-retaliation" provisions as overbroad. Plaintiffs’ proposal to require some further disclosures to advertisers was rejected. The Court did agree with a provision under which Google will be required to disclose material changes to its ad auction rules to promote greater transparency and to prevent Google from increasing prices to advertisers by secretly "fine-tuning" its ad auction procedures.²²

THE COURT REJECTS PROHIBITIONS ON "SELF-PREFERENCING"

A major topic of discussion among antitrust practitioners is whether "self-preferencing" by dominant "Big Tech" platforms- i.e., favoring or advantaging its own software services as opposed to those offered by rivals- should be deemed to be "exclusionary" conduct.

Plaintiffs had proposed provisions to restrict Google from engaging in a wide variety of "self-preferencing" behavior.²³ Judge Mehta rejected these provisions "for reason both legal and factual". First, he ruled that that such behavior was not of the same type or class of illegal conduct which was the subject of this case. Second, the proposed ban on "self-preferencing" "goes too far in that it would hamstring Google’s ability to compete."²⁴ This is especially so since other companies in the GenAI space leverage their own products to distribute their GenAI technologies- Meta delivers its GenAI models through Instagram and WhatsApp; xAI makes Grok available through X; Microsoft had integrated Copilot into Edge and Bing. As Judge Mehta states, "The court will not hobble Google’s competitiveness by prohibiting self-preferencing of its own GenAI technologies, when that is precisely how the emerging-and highly competitive- GenAI marketplace operates."²⁵

TERM OF FINAL JUDGMENT

Rather than accept the Plaintiffs proposal that the Final Judgment last ten (10) years, the Court ruled that a six (6) year term is appropriate. This contemplates that it will take one year to establish the Technical Committee and the processes for implementation of the Judgment, and then the Judgment would be in force for five years (not counting time for appeal). Judge Mehta noted, "Technological advancement and product differentiation are what will change market dynamics, not a decade-long judicial decree."²⁶ But his opinion does leave open the possibility of continued judicial oversight and potential changes to his remedy. For example, in a footnote, he wonders about the possibility that Google would only be allowed "to make payments for distribution but not for default distribution." He then states that "[w]ithout a more substantial record" on this possibility, "the court is

[Page 6]

presently ill-equipped to consider it."²⁷ It is not clear whether he is envisioning some later proceeding on this possibility.

CONCLUSIONS, QUESTIONS, AND LESSONS

The ink is barely dry on the decision but, five years into this historic case, we can draw some conclusions and ask some questions.

First, the antitrust fight against Google is probably not even at halftime. There will be disputes over the final language of the Final Judgment; there will be appeals of Judge Mehta’s opinion—perhaps from both Google and the United States; and there will be six years of antitrust oversight over Google, by the Technical Committee and the Court. And there soon will be a remedies opinion in the Google AdTech case. It is quite possible that the Google cases, like the Microsoft cases of the 1990s, will play out over a decade. Just as each historic Microsoft opinion begat more historic proceedings and opinions, the 2020s Google cases (so far, Search, Ad Tech, and Play Store) might lead to new cases, and Google roman numerals might be in the future.

Second, we will know whether Judge Mehta’s decision went far enough when we access whatever "AI/Search" is in six years. Will an overwhelming majority still access the world’s information through a Google property? Will there be new companies offering access points? Will some new company challenge Google in the wake of Google’s antitrust troubles, just as Google challenged Microsoft in the wake of Microsoft’s antitrust cases? This said, perhaps technology will have advanced in some way that the Microsoft/Google framework no longer governs. Both cases assumed users would access a single, universal source of information (the internet) through a single gate. It might be that concepts like desktops and search engines will soon be as quaint as a DOS prompt.

Third, the most important words in the judgment might be "qualified competitor." These companies will be able to benefit from the data-sharing and syndication remedies. Judge Mehta made clear that the definition of "qualified competitor" should include companies offering GenAI products who plan to compete "with" general search engines.²⁸ The court, in line with the point above, imagined a future where some form of AI renders "Search" obsolete. He wrote that both AI and Search have the "capacity to fulfill a broad array of informational needs."²⁹ So, while the case was about Google’s "10 blue links," the remedy decision imagines some other future—one beyond the ken of mere lawyers.

Fourth, company reputation matters a lot. In several places, Judge Mehta emphasizes the court’s view that Google’s monopoly was in large part attributable to its superior product (and that the superior product was in large part attributable to Google’s innovation, not the power of defaults). "After two complete trials, this court cannot find that Google’s market dominance is sufficiently attributable to its illegal conduct to justify divestiture."³⁰ He also repeated a passage from the liability opinion concluding that "Google has not achieved market dominance by happenstance. It has hired thousands of highly skilled

[Page 7]

engineers, innovated consistently and made shrewd business decisions."³¹ Contrast these views with the fact that in Microsoft, there was little question that Netscape had the superior product. Netscape was "the new competitor ‘born’ on the internet" that threatened Microsoft’s entire business model. Here, Judge Mehta appears to see Google as more of a 1990s Netscape (particularly given looming AI competition) than a 1990s Microsoft, although today’s Microsoft would disagree.

Fifth, the timing of government enforcement matters. The United States filed the Google Search case in October 2020 but the government could have filed it years earlier: seven years earlier, the FTC had declined to bring an enforcement action against Google, and the payments for defaults had been in place for years. Perhaps if the case had been filed earlier, Judge Mehta’s remedies decision would have been different.

One of the driving forces for his caution—to the frustration of some—is that the status quo of defaults was so strong that it would be dangerous to disrupt it. He wrote that "Plaintiffs acknowledge the possibility of adverse market effects from a complete payment ban but implore to focus on the task of restoring competition to the relevant product markets. They believe that although there may be short-term harm to some market actors, they will benefit in the long run from increased competition. Acting in equity, however the court cannot be so myopic. It must consider the harms that might befall other market actors, even if that means, as here, forgoing a remedy that could help restore competition."³²

Judge Mehta also seems to have been influenced by a limited runway left for Search. Given the emergence of GenAI, he also understood that the market dynamics have changed since the case was originally filed and even after the liability decision a year ago. So Judge Mehta might have thought an additional reason not to disrupt Search was that the harm had already been done and strong relief would not make a difference because of new technologies.

Imagine a Google Search case filed in 2013 or 2016, the status quo would have been less entrenched and there would have been no AI competition even on the horizon. Or imagine a Google Search case resolved in three or four years, rather than five years. Even a one year difference would have meant the AI threat would have been less apparent.

* * *

Judge Mehta carefully and meticulously considered the remedy proposals. He approached his task with humility and caution, explaining that he does not want to be the regulator of the software industry over the next six years. Some would use different words than humility and caution, exasperated that the first part of the antitrust revolution ended with a whimper. The jury will be out for several years for the final verdict on his decision.

[Page 8]

——–

Notes:

^1. Samuel Miller was the lead prosecutor for the United States in the licensing case against Microsoft. Ryan Sandrock worked on the early stages of the Google Search case during his time at the Antitrust Division. Sam and Ryan together teach Antitrust and Intellectual Property at the University of California College of the Law—San Francisco. The views expressed in this article are solely those of the authors.

^2. See Ford Motor Co. v United States, 405 U.S. 562,573 (1972); United States v. Microsoft Corp., 253 F. 3d 34,103 (D.C. Cir. 2001) (en banc) (hereinafter Microsoft).

^3. Microsoft, 253 F. 3d at 107.

^4. United States v. Google LLC, 747 F. Supp. 3d 1, 159-163 (D.D.C. 2024) (hereinafter Google Liability Opn.); United States v. Google LLC, case 1:20-cv-03010, Dkt. 1436 at p. 149-150, 2025 U.S. Dist. LEXIS 170459 (D.D. C. Sept. 2, 2025) (hereinafter Google Remedy Opn.) (page citations in this article will be to the Slip Opinion).

^5. Google Remedy Opn. at pp. 40-41.

^6. Google Remedy Opn. at pp. 104-105.

^7. See Final Judgment in U.S. v. Microsoft Corp., case 94-1561, entered Aug. 21, 1995, 1995 U.S, Dist. LEXIS 20533 (D.D. C. Aug. 21, 1995).

^8. Google Remedy Opn. at p. 108.

^9. Google Remedy Opn. at p. 108.

^10. Google Remedy Opn. at pp. 114-115.

^11. Google Remedy Opn. at p. 119.

^12. Google remedy Opn. at p. 188-190.

^13. Google was paying over $20 billion per year to distributors such as Apple as revenue share or for default placement. Google Remedy Opn. at p. 119.

^14. Google Remedy Opn. at pp. 121-125.

^15. Google Remedy Opn. at pp. 127-128.

^16. Erik Hovenkamp (@ErikHovenkamp), X (September 5, 2025)

^17. Google Remedy Opn. at p. 130.

^18. Google Remedy Opn. at p. 130, discussing In re Google Play Store, 2025 WL 2167402, at *16-17 and Massachusetts, 373 F. 3d at 1216-25.

^19. Google Remedy Opn. at pp. 128-146.

^20. Google Remedy Opn. at pp. 162-163.

^21. Google Remedy Opn. at 9.

^22. In a separate case pending in the Eastern District of Virginia, Google was found to have violated Section 2 with respect to certain markets relating to ad technologies, and is awaiting a remedy decision in that case, which may affect rules for ad auctions.

^23. Google Remedy Opn. at p. 215.

^24. Google Remedy Opn. at p. 217.

^25. Google Remedy Opn. at pp. 217-218.

^26. Google Remedy Opn. at p. 221.

^27. Google Remedy Opn. at p. 128, n. 14.

^28. Google Remedy Opn. at p. 103.

^29. Google Remedy Opn. at pp. 99-100.

^30. Google Remedy Opn. at p. 221.

^31. Google Remedy Opn. at p. 77.

^32. Google Remedy Opn. at pp. 125-126.