The Remedy Is a Two-Sided Opening
The Epic Games v. Google remedy is best understood as a redesign of relationships inside Android, not as a single command to lower a fee. The Ninth Circuit affirmed liability findings about both app distribution and in-app billing, then upheld an injunction that restricts some of Google’s contracts while requiring new routes for rival stores and developers. The useful judgment is to separate what Google is forbidden to do from what it must make possible, and then connect each change to the bottleneck it addresses.
Liability tells us why relief was needed
A liability finding answers whether the defendant violated the law under the facts proved at trial. Equitable relief is the court’s forward-looking response: an order meant to stop unlawful conduct, prevent its recurrence, and repair its consequences. In this case, the jury found that Epic had proven relevant markets for Android app distribution and Android in-app billing. It found that Google had willfully acquired or maintained monopoly power, unreasonably restrained trade, and unlawfully tied use of the Play Store to Google Play Billing. The Ninth Circuit affirmed that verdict.
That verdict did not itself specify a new app-store architecture. It established the legal basis for the district court to craft a remedy. Under Section 16 of the Clayton Act, a private party can seek an injunction against threatened loss or damage from an antitrust violation. The appellate court explained that this power is not limited to telling a defendant to stop the exact conduct that produced the verdict. A court may also deny the defendant the benefits of the violation and address practices likely to produce monopolization again, as long as the relief has a significant causal connection to the violation and is a reasonable way to eliminate its consequences.
That distinction matters when interpreting the word “open.” The injunction does not say that every developer may use every payment processor, that every store must be accepted without review, or that Google must give away every platform service. It identifies particular levers that helped protect Play and then changes the conditions around those levers.
The order closes several routes to exclusion
The first component is prohibitory. The injunction stops Google from using certain arrangements to insulate Play from distribution competition. As described by the Ninth Circuit, Google may not share Play revenue with actual or prospective entrants in ways that protect Play from a rival. It may not condition payment or access to Play on an app launching first or exclusively there. It may not condition an OEM’s Play placement on an agreement not to preinstall another app store in a particular location on an Android phone. And it may not continue requiring Google Play Billing for all apps distributed through Play.
These provisions are not an attempt to regulate every commercial term in the Android ecosystem. They track the conduct the trial record connected to Play’s position. The court had heard evidence about agreements that discouraged OEMs from carrying rival stores and about arrangements that discouraged developers from launching on other Android stores. The remedy therefore targets exclusivity, placement, and payment conditions as mechanisms of foreclosure rather than treating “Google” or “Play” as the problem in the abstract.
The order’s billing provision is particularly easy to overread. Prohibiting a blanket requirement to use Play Billing is not the same as abolishing Play Billing. Google can still operate its billing service, and a developer can still use it. The change is that Play cannot make its distribution channel depend universally on that one billing route. The operational question—what an eligible developer must do to use another route—belongs to the program rules, not to the liability verdict alone.
The affirmative remedy repairs both sides of the market
Stopping exclusionary arrangements would remove some barriers, but it would not automatically undo the network effects that those barriers helped strengthen. A rival store could still have too few apps to attract users, and too few users to attract developers. That is why the injunction also contains affirmative access provisions.
The catalog-access remedy works on the supply side. It requires Google to permit third-party Android app stores to access the Google Play catalog so that they can offer Play apps to their users. For downloads processed by Google Play, users must be able to complete downloads of apps that are available only on Play on the same terms as if they downloaded them directly from Play. Developers must also have a mechanism to opt out of inclusion in catalog access for a particular third-party store.
The app-store-distribution remedy works on the user side. Google may not prohibit the distribution of third-party Android app distribution platforms or stores through the Play Store. This gives a rival a path to users who are already accustomed to obtaining software through Play. The court described app stores as a type of app for this purpose: if Google can carry other third-party software using technical criteria, it cannot simply exclude rival stores as a matter of policy while claiming that the remedy requires a wholly different kind of service.
The two provisions are separate because they solve different problems. Catalog access answers, “Can a rival store offer enough software to matter?” App-store distribution answers, “Can that rival get in front of Android users who currently rely on Play?” Neither guarantee is a promise that a rival will succeed. Together, however, they change the starting conditions under which success can be attempted.
The market relationship in the diagram is the point. Google Play remains a central intermediary, but the injunction changes who can connect to its catalog, who can be distributed through its platform, and how an app’s transaction path may be configured. A new route matters only if it crosses the interface where the old barrier operated.
Why the court allowed a form of compelled access
Google argued that catalog access imposed an impermissible duty to deal—an obligation to share with competitors that antitrust law does not ordinarily create. The Ninth Circuit drew a critical line. A company’s unilateral refusal to deal may be lawful in some circumstances when the question is whether that refusal itself violates Section 2 of the Sherman Act. That does not mean a court loses the power, after finding an antitrust violation, to require conduct that is necessary to unwind the violation’s consequences.
The opinion relies on earlier cases in which courts required defendants to deal with rivals or customers on specified terms as part of an antitrust remedy. The relevant question was not whether Google would have chosen to offer catalog access as a new commercial product. It was whether requiring access was a reasonable method of counteracting the network effects that Google’s exclusionary conduct had unfairly enhanced.
The causation standard was narrower than Google’s objection suggested. The district court did not have to prove that every part of Play’s network advantage came from unlawful conduct. It had to find a significant connection between the conduct addressed by the remedy—Google’s technical and contractual exclusion of other app-store developers—and the monopoly that the jury found. The record supported the finding that Google had erected barriers to protect the network effects it was using to maintain dominance. Temporarily lowering those barriers therefore addressed a consequence of the proven violation.
This is why the remedy is not a general rule that every platform must share every asset with every competitor. Its legal justification depends on the market, the conduct found unlawful, and the relationship between the remedy and the harm. The same reasoning also explains the three-year catalog-access window. The court treated it as a discrete period in which rival stores could obtain a fair chance to establish themselves with a competitive library, not as a permanent transfer of Google’s accumulated catalog.
Opening does not remove safety or price controls
The affirmative provisions preserve a platform role for Google. The injunction allows Google to take reasonable measures to ensure that stores and the apps they offer are safe from a computer-systems and security standpoint, do not offer illegal goods or services, and comply with Google’s content standards. It also permits Google to charge a reasonable fee based on its actual costs for those measures.
That qualification is not an incidental footnote. An app store is a distribution platform for software, and software can create security, privacy, intellectual-property, and consumer-protection problems. The court’s answer was not to treat those concerns as an automatic veto. It required the security measures to be reasonable and, for rival stores, comparable to the measures Google already applies to apps on Play. The pricing rule also matters. A merely nondiscriminatory fee could still exclude every rival if Google set the same excessive price for all of them. A reasonable, cost-based fee leaves Google room to recover legitimate review costs while limiting the use of price as a substitute for an outright ban.
The appellate opinion records disagreement about the longer-term security implications. Some participants warned that the order could lower barriers for malicious third parties; others argued that Google’s existing restrictions were not necessary for security and that competition could produce better long-term protection. The court upheld the safeguards in the injunction, but that holding is not an empirical finding that the resulting ecosystem will be safer. The selected evidence does not measure post-implementation security outcomes.
A legal opening still needs an implementation process
The court anticipated that the most difficult disputes would involve technical choices rather than the broad meaning of the order. It created a three-person Technical Committee with one member selected by Epic, one by Google, and a third selected by those representatives. If the committee cannot resolve a dispute, a party may ask the district court to decide. Google bears the burden of showing that a disputed technical requirement is strictly necessary for the safety and security of users and developers. The committee may recommend whether to extend a deadline, but it cannot extend one on its own.
This arrangement does not replace the court’s authority. It supplies specialized help while leaving unresolved legal and remedial questions with the judge. The Ninth Circuit held that the injunction gave Google fair notice of what it had to do without spelling out every metadata field, refresh interval, or implementation recipe. An order can be specific about the required result and the constraints on compliance without becoming a cookbook for the defendant.
The timing reinforces the same point. The injunction described in the opinion began on November 1, 2024, ran for three years to November 1, 2027, and gave Google eight months to develop the technology needed for catalog access and app-store distribution. A footnote records that, at the time of the appeal, only the provision concerning payments to manufacturers not to preinstall Play competitors had taken effect while the other provisions were stayed pending appeal. The Ninth Circuit ultimately affirmed the injunction in full.
The remedy therefore has two dimensions. It removes specific tools of exclusion, and it creates supervised interfaces through which alternatives can become viable. The next question is whether a reader can use that structure to distinguish catalog access from rival-store distribution, and distribution openings from the separate choices around billing and external links.