Europe keeps asking where its technology champions are. In Google Android, the Court of Justice of the European Union (CJEU) offered part of the answer: build a successful platform, and Brussels may spend the next decade treating its architecture as evidence.
The CJEU’s final judgment in Google Android, handed down last week, will be celebrated in Brussels as a triumph of public enforcement over Big Tech. It deserves a less triumphant reading.
The judgment ends an eight-year legal fight by leaving Google and Alphabet with a fine of roughly €4.125 billion for contractual practices tied to Android, Google Search, Chrome, and the Play Store. The court accepted that Google abused its dominance by using Android distribution terms, preinstallation conditions, anti-fragmentation obligations, and related arrangements to favor its own search and browser products.
The fine is painful. The precedent is worse.
The CJEU approved important parts of the General Court’s analysis. It allowed courts to consider economic context without systematically constructing a counterfactual. It also confirmed that liability does not always depend on proof that the practices could foreclose an “as-efficient competitor” (AEC). That test asks whether a rival as efficient as the dominant firm could compete despite the challenged conduct.
That doctrinal signal matters more than the penalty. A €4 billion fine stings. But the larger cost comes from the precedent’s effects on platform design, investment incentives, and the legal expectations facing future European technology firms. If Article 102 of the Treaty on the Functioning of the European Union (TFEU)—which governs abuse of dominance—condemns ordinary platform governance whenever rivals dislike the outcome, Europe will not get more innovation. It will get more litigation, more regulatory redesign of products, and fewer firms willing to build integrated platforms in the first place.
The Trouble With Treating Scale as Suspicious
This is an odd place for Europe to be. Mario Draghi’s report on European competitiveness warned that Europe largely missed the internet-led digital revolution. It also found that the productivity gap between the European Union and the United States came largely from the technology sector, and that only four of the world’s top 50 technology companies are European.
When it comes to closing the innovation gap with the United States and China, the numbers are grim. No EU company with a market capitalization above €100 billion has been built from scratch in the past 50 years. European firms spent €270 billion less on research and innovation than their U.S. counterparts in 2021. Nearly 30% of European-founded unicorns relocated abroad between 2008 and 2021, overwhelmingly to the United States.
Alas, the Google Android approach points in precisely the wrong direction. Europe needs more risk-taking, not more legal suspicion of business success. It needs more scale, not more hostility to the contractual tools that help platforms coordinate device makers, app developers, consumers, advertisers, and service providers. It needs more experimentation in monetization, default design, distribution, and integration, not more efforts to recast successful platform architecture as anticompetitive conduct after the fact.
Android is not a conventional product sold in isolation. It is a mobile operating system that links handset makers, app developers, carriers, consumers, advertisers, browser developers, search providers, and Google itself. Google made Android broadly available, limited fragmentation, supported developers, maintained compatibility, and monetized much of the project through complementary services such as search.
That model can be criticized, but it is not obviously anticompetitive. Even if it were, the central law & economics question remains: compared with what?
Compared with a world in which Google could not assure prominent distribution for Search and Chrome, would Android have been offered as broadly, cheaply, and consistently? Would handset makers have received the same free or low-cost operating system? Would developers have enjoyed the same compatibility and reach? Would consumers have received lower prices, better app availability, safer devices, and stronger competition against Apple’s vertically integrated iOS model?
Those questions are not side issues. They are the core of the case. A platform’s defaults, bundles, and compatibility rules often are not dirty tricks. They are part of the machinery that allows the platform to exist.
International Center for Law & Economics (ICLE) scholar Dirk Auer made this point years ago. His 2020 analysis of the European Commission’s Android decision argued that the Commission’s factual account did not prove that Google harmed competition or consumers. He criticized the decision’s market definition, dominance analysis, treatment of Apple, theories of harm, and innovation claims.
Two years earlier, Auer warned that the Android decision threatened the viability of Android’s open-source model by meddling with the governance rules through which Google competed with Apple. One need not accept every detail of those critiques to see that the core warning has aged well. Competition law becomes dangerous when it treats platform coordination as suspect merely because rivals would prefer different terms.
Geoffrey Manne, Lazar Radic, Dirk Auer, and other ICLE authors have made the same broader point. Digital competition rules often dress up competitor protection as competition policy. Their critique is not that digital platforms can do no wrong. It is that many recent rules and decisions prioritize fairness, contestability, and redistribution among firms over efficiency, innovation, and consumer welfare.
That approach can shield less efficient rivals, dilute the advantages successful platforms earned, and make firms less willing to invest in the next platform. Google Android shows the problem. A rival search engine may want guaranteed placement. A rival browser may want default status. A device maker may want Google’s app store without Google’s compatibility rules. Those are understandable business preferences. They are not evidence that consumers were harmed.
Defaults Are Design, Not a Smoking Gun
The Android judgment is part of a larger project. It belongs with Google Shopping (2024) and Europe’s broader campaign against self-preferencing, platform integration, and so-called gatekeeper advantages.
In Google Shopping, the CJEU upheld an expansive theory of leveraging through the design and display of search results. The court also confirmed that the Commission did not need to reconstruct a full counterfactual or show foreclosure of an as-efficient competitor in the way a more disciplined effects-based inquiry might require.
The common thread is not a traditional showing that consumers paid more, output fell, or innovation slowed. It is the intuition that a successful platform’s control over valuable distribution is suspect when rivals would prefer another allocation. That intuition shifts antitrust away from consumer welfare and toward platform-neutrality regulation. It treats search results, defaults, app-store licenses, and product design as quasi-public utilities to be allocated by administrative judgment.
High-technology competition rarely works that way. Firms invest in platforms partly because integration can yield advantages. If those advantages later become duties to assist rivals, the incentive to build the next platform weakens before anyone writes the first line of code.
The old antitrust maxim remains indispensable. Competition law should protect competition, not competitors. In digital markets, that maxim must account for dynamic competition. If a firm invests in a platform, attracts developers, subsidizes users, solves compatibility problems, and offers a product consumers value, rivals will often lose ground. That is not a pathology. It is competition among business models. Some are integrated, some are modular, some are ad-funded, and some are subscription-funded.
Defaults illustrate the problem. The Android decision gives great weight to status quo bias, meaning the tendency of consumers to stick with what comes preinstalled. That tendency is real, but it does not prove anticompetitive harm.
Defaults reduce search costs. They make a device usable out of the box. They can signal quality. They can fund a zero-price operating system. Users can change them. Rivals can counter them through brand, distribution, and product quality. To say that a default matters is not to say that it excludes competition. The relevant question is whether the default raises quality-adjusted prices, reduces output, degrades quality, suppresses innovation, or prevents equally efficient rivals from reaching consumers through realistic channels.
A consumer-welfare-oriented court would insist on a rigorous counterfactual. Without one, enforcement can confuse the observed success of an integrated strategy with proof that the strategy caused unlawful foreclosure.
That matters even more in platform markets, because an intervention on one side can harm participants on another. Requiring Google to unbundle, weaken defaults, or tolerate fragmentation may help rival search engines or browsers. It may also reduce Google’s incentive to invest in Android, increase device makers’ costs, reduce security, complicate developer support, or make Android less attractive against Apple. A serious counterfactual would compare those possibilities. The Android judgment’s tolerance for avoiding that analysis invites false positives.
The court’s treatment of the as-efficient-competitor principle raises the same concern. The AEC test is not a talisman, and it will not fit every nonprice case. But the principle behind it is vital. Article 102 should not preserve rivals who lose because they are less attractive to consumers, less innovative, less efficient, or slower to develop alternative distribution channels. It should intervene when dominant-firm conduct prevents competition on the merits from working.
A doctrine that loosens the AEC inquiry because the market is digital risks turning the special features of technology markets into an excuse for weaker proof.
Article 102 Needs a Consumer-Welfare Spine
The Android judgment departs from better instincts in recent European case law.
In Intel (2017), the CJEU required careful attention to evidence and economic analysis when a dominant firm argued that its rebates could not foreclose as-efficient competitors. In Servizio Elettrico Nazionale (2022), the court tied Article 102 to practices that may harm consumers, including indirectly through harm to the competitive process. It also recognized that a dominant firm may justify its conduct by showing positive consumer effects. In Unilever Italia (2023), the court held that exclusivity clauses must be capable of exclusionary effects and that authorities must assess evidence submitted by the dominant undertaking, including economic studies and AEC-type evidence where relevant.
Those decisions do not create a U.S.-style antitrust code. Europe need not copy American doctrine wholesale. But they display virtues European competition policy badly needs: limiting principles, attention to effects, seriousness about evidence, and a willingness to distinguish exclusion caused by anticompetitive conduct from exclusion caused by superior performance.
ICLE’s 2024 comments on the Commission’s draft Article 102 guidelines describe this more effects-based line of cases as including Intel, Servizio Elettrico Nazionale, Unilever Italia, and Intel Renvoi. The comments also criticize the Commission’s more formalistic turn, which risks chilling procompetitive conduct when Europe already trails in productivity and competitiveness.
In 2024, I argued that Europe’s latest monopolization-policy direction threatens innovative business practices that promote high-tech growth. I criticized presumptions against conduct such as self-preferencing, as Brian Albrecht and Manne also discussed in 2025, along with exclusivity, tying, low pricing, and other practices that may benefit consumers.
The institutional problem is simple. Dominant firms facing vague standards, hard-to-prove efficiency defenses, and massive fines will pull their punches. They will avoid aggressive product improvement, integration, and efficient contracting because the penalty for guessing wrong is enormous.
That is the real danger of Android. The judgment tells large platforms, and aspiring European platforms, that product integration and contractual coordination may later be judged through the lens of rival disadvantage. It tells firms that a successful default can become evidence of status quo bias, monetization tied to distribution can become unlawful leveraging, and compatibility obligations can become obstruction of alternatives. It tells entrepreneurs that the reward for scale may be a decade of litigation and a court-approved redesign of the business model.
Google should not receive immunity because it innovated. Dominance does not create a license to deceive, coerce, sabotage interoperability, or sacrifice product quality merely to block rivals. Nor is every default, bundle, or exclusive arrangement benign. Some practices may exclude equally efficient rivals and harm consumers. But the legal system must prove that harm rather than infer it from the fact that rivals lost a preferred path to users.
The distinction matters because innovation is path-dependent. Firms do not merely respond to today’s fine. They respond to the legal environment they expect over the next decade. If that environment tells them that successful platform governance will be second-guessed, rational firms will design less ambitiously. They will avoid business models that require cross-subsidy. They will offer fewer integrated features. They will litigate before launching. They will negotiate with regulators before testing products with consumers.
None of that appears in a static foreclosure chart. All of it matters for dynamic competition.
The predictable beneficiaries are not necessarily consumers. They are often business users, rivals, and follow-on plaintiffs who can turn a public judgment into private damages claims. Once a court validates a broad theory of exclusion, litigation incentives change. Rivals can seek compensation for lost distribution, not because consumers paid more or received worse products, but because the rival’s path to scale was harder. Competition law then becomes a tool for redistributing returns from platform investment. It stops disciplining harm to competition and starts refereeing disappointment.
The Next Platform Will Read the Fine Print
A wiser Article 102 framework for platform cases would begin with consumer welfare and dynamic competition, not an abstract preference for rival access.
First, courts should require a serious counterfactual. That inquiry should consider the platform’s monetization model, investment incentives, product quality, security, compatibility, and competition among platforms.
Second, when a dominant firm submits plausible evidence that as-efficient rivals could compete, or that the practice creates efficiencies, authorities should engage that evidence rather than dismiss it as unnecessary.
Third, integration, tying, preinstallation, anti-fragmentation rules, and self-preferencing should not be treated as suspect labels. They are business practices. They may help consumers or harm them depending on context.
Fourth, courts should distinguish access to a rival’s created asset from suppression of competition itself. Google’s rivals may want access to Android distribution on terms that maximize their reach. But Android’s distribution architecture was not a natural resource. Google built and maintained it through investment, governance, licensing, developer support, and monetization. Antitrust should hesitate before converting such assets into regulated opportunities for rivals. Forcing a platform to subsidize competitors may sound like fairness. It can also reduce the reward for creating the next platform.
European courts should also remember Draghi’s warning. Europe is not overrun with homegrown global technology platforms. It is struggling to create them. A competition policy that treats scale, integration, default design, and platform control as presumptively suspicious will not close that gap. It will widen it. Europe cannot regulate its way into technological leadership by making the business models of technological leadership legally precarious.
This most recent Android judgment is a setback because the court’s reasoning risks pulling European abuse-of-dominance law away from the economically grounded path suggested by Intel, Servizio Elettrico Nazionale, and Unilever Italia. Future CJEU decisions should return to that path. They should require evidence, counterfactuals, efficiency analysis, and a clear showing of harm to consumers and dynamic competition.
Above all, they should reject the easy slide from harm to businesses to harm to competition. Europe’s greatest high-tech competition problem is not that successful platforms innovate too much. It is that too few firms believe Europe will reward them if they build the next one.
