Google runs out of road as EU top court locks in the €4.1 billion Android fine

On 2 July 2026, the Court of Justice of the European Union dismissed the appeal brought by Google and its parent company Alphabet in case C-738/22 P, confirming a fine of €4.125 billion for anticompetitive practices tied to the Android operating system. The judgment is final. There is no higher court, no further appeal, no procedural mechanism left through which Google can reopen the case within the EU legal order. A dispute that began with a formal Commission decision in July 2018, and with an investigation that stretches back even further, has ended with a complete defeat for one of the most powerful companies in the world.

The ruling that ends an eight-year fight

The core of the case was never Android itself. Android, as an open-source project, was not on trial. What the European Commission attacked in 2018, and what two levels of EU courts have now confirmed, was the web of contracts Google built around Android. Those contracts required smartphone manufacturers who wanted the Google Play Store to also pre-install Google Search and the Chrome browser. They barred those same manufacturers from selling any device running an unapproved variant of Android, the so-called forks. And for a period, they paid manufacturers and mobile network operators for exclusive pre-installation of Google Search. The Commission concluded that this architecture existed for one purpose: to protect and extend the dominance of Google Search on mobile devices at the exact moment the world was shifting from desktop computers to smartphones.

The Court of Justice agreed with that reading in almost every respect that mattered. In its press communication, the court stated plainly that the appeal was dismissed, “thereby confirming the penalty imposed for Google Search’s abuse of a dominant position in the context of the Android operating system.” The Second Chamber found no error of law in the General Court’s 2022 judgment, which had itself upheld the bulk of the Commission’s 2018 decision while trimming the fine from €4.343 billion to €4.125 billion.

For readers outside competition law, the significance of finality deserves emphasis. EU antitrust decisions live in a state of suspension for as long as appeals run. A company under appeal can tell its investors, its partners and the public that the matter remains contested, that the courts may yet see things differently, that the fine is provisional. Google used that framing for eight years. As of 2 July 2026, that framing is gone. The finding that Google abused a dominant position between 2011 and 2018 is now settled law across all 27 member states, binding on every national court in the Union.

That has three immediate consequences. First, the money. Google had already provisioned for the fine, and companies typically pay disputed EU fines into escrow or provide bank guarantees while appeals run, so the cash impact on Alphabet is administrative rather than dramatic. Alphabet shares dipped roughly one percent in premarket trading after the ruling, a market reaction that says investors had long priced in this outcome. Second, the precedent. The judgment settles contested doctrinal questions about tying, pre-installation, default effects and the evidentiary burden the Commission must meet, and those answers will shape every platform case Brussels brings from now on. Third, the exposure. A final finding of abuse is a gift to private litigants. Any company that can plausibly argue it was harmed by Google’s Android contracts between 2011 and 2018 can now walk into a national court with the liability question already answered and argue only about damages.

The timing adds its own weight. The ruling lands in a period when the EU has moved beyond case-by-case antitrust enforcement into the era of the Digital Markets Act, which imposes standing obligations on designated gatekeepers rather than punishing abuse after the fact. Google is a designated gatekeeper for Android, Search, Chrome, Google Play and other services. Earlier in 2026, the company was ordered to dismantle technical barriers that kept rival AI search assistants from operating freely on Android and to share certain search data with competitors. The Android judgment does not create those obligations, but it validates the worldview behind them: that Google’s control over mobile distribution channels is a competition problem serious enough to justify structural intervention.

There is also a human arc to the story worth recording. The case was built under Margrethe Vestager, the Danish competition commissioner who made confronting Silicon Valley the defining project of her decade in office. She left the Commission in 2024 with her three great Google cases — Shopping, Android and AdSense — in various states of judicial review. The Shopping fine of €2.4 billion was confirmed by the Court of Justice in September 2024. Android, the largest of the three, is now confirmed as well. Whatever one thinks of Vestager’s approach, the courts have now told her successors that the legal theory underneath it was sound.

The rest of this analysis takes the case apart piece by piece: the contracts, the economics, the legal doctrine, the eight-year procedural history, and the practical consequences for manufacturers, developers, advertisers, search marketers and every business that lives downstream of Google’s mobile ecosystem.

A €4.1 billion penalty and the numbers behind it

The figure most headlines carry, €4.1 billion, is a rounding of the precise amount: €4,125,000,000, imposed on Google LLC, with Alphabet Inc. jointly and severally liable for €1,921,666,000 of it. The original Commission fine of 18 July 2018 was €4,342,865,000. The General Court reduced it in September 2022 after annulling one part of the decision, the part concerning certain portfolio-based revenue share agreements, and after taking a slightly different view of the gravity of the conduct. The Court of Justice has now confirmed the revised amount without further adjustment.

The methodology behind the number follows the Commission’s 2006 fining guidelines. The Commission took the value of Google’s relevant sales in the European Economic Area during 2017, the last full year of the infringement, applied a gravity coefficient of 11 percent, multiplied the result by the duration of the infringement, roughly 7.52 years, and added an additional deterrent amount equal to 11 percent of the 2017 sales value. The relevant sales here were, in substance, search advertising revenues earned through the channels the abusive conduct protected. That is why the fine is so large: Google’s European search advertising business is enormous, and the infringement ran for the better part of a decade.

At the time it was imposed, the Android fine was the largest antitrust penalty ever issued by any competition authority anywhere in the world, comfortably exceeding the €2.42 billion Google Shopping fine of 2017, which had itself broken the previous record held by Intel. It remains the largest fine the Commission has ever collected in a single abuse-of-dominance case. Across the Shopping, Android and AdSense decisions, plus subsequent enforcement, Google has accumulated close to €11 billion in EU antitrust fines over the past decade, a figure without parallel among technology companies.

Context matters when judging whether such a number bites. Alphabet’s annual net profit has run above $100 billion in recent years, which means the confirmed Android fine represents less than three percent of a single year’s profit. Critics of EU enforcement point to this ratio as evidence that even record fines function as a cost of doing business for firms of Google’s scale. Defenders reply that the fine was never the main event. The remedies were. The 2018 decision required Google to end the illegal conduct within 90 days, and it was the changes to Google’s contracts — unbundling Chrome and Search from the Play Store licence in Europe, dropping exclusivity conditions, eventually introducing a choice screen for search engines on Android devices — that altered the market, not the cheque.

There is a second financial dimension that the headline number hides: interest and litigation exposure. Companies that pay EU fines provisionally and later win appeals get their money back with interest; companies that lose, as Google now definitively has, simply lose. More important is what lawyers call follow-on exposure. A final infringement decision is binding proof of illegality in damages actions before national courts across the EU. Sweden offered a preview just days before the CJEU ruling, when a Swedish court ordered Google to pay roughly $1.5 billion in damages to PriceRunner, the comparison-shopping service now owned by Klarna, in litigation that flowed from the separate 2017 Shopping decision. The Android judgment opens a comparable door for a different set of claimants: rival search engines, browser makers, app store operators and device manufacturers who can argue the contracts foreclosed them.

Key figures in the Android case

The table condenses the financial skeleton of the case. Two things stand out from it: the courts changed the number only modestly across eight years of litigation, and the fine, however historic in absolute terms, is small relative to the profit pool the protected conduct helped generate. Both facts feed directly into the policy debate about whether fines alone can discipline dominant platforms, a debate the EU has effectively answered by legislating the Digital Markets Act.

One more number deserves attention: the duration of the proceedings themselves. From the Commission’s statement of objections in April 2016 to final judgment in July 2026 is more than a decade. From the opening of the formal investigation in 2015, longer still. During that entire period the mobile market kept moving, AI assistants arrived, and Google’s contracts were amended twice over. The gap between the speed of digital markets and the speed of judicial review is now one of the strongest arguments regulators make for ex-ante rules, and the Android case will be cited as its leading exhibit.

Google Android as a business model, explained

To understand the case, start with the deal Google offered the mobile industry in the late 2000s. Android would be free. Any manufacturer could take the open-source code, build a phone around it, and pay Google nothing. Apple charged for integration through its closed ecosystem; Microsoft charged licence fees for Windows Mobile. Google gave the operating system away, and it did so because the operating system was never the product. The product was the user’s attention, and the revenue was search advertising.

Google’s economics rest overwhelmingly on advertising sold against search queries. Every search a user runs on Google is an opportunity to show ads; every search run elsewhere is revenue lost. In the desktop era, Google secured its query flow by paying browser makers for default placement and by making a search product good enough that users typed google.com voluntarily. The smartphone changed the physics of that arrangement. On a phone, users do not configure their software. They use what is on the home screen out of the box. Whoever controls the pre-installed search widget, the pre-installed browser and its default search engine effectively controls the query stream from that device.

Android was Google’s answer to the risk that someone else — Apple, Microsoft, or a carrier — would control that stream. But open-sourcing the OS created a new risk: if anyone could take Android and strip Google out of it, Amazon or Samsung or a Chinese vendor could ship millions of “Android” phones that sent queries elsewhere. Google managed this tension through a layered contractual structure sitting on top of the open-source code, and it is this structure the EU condemned.

The first layer was the Mobile Application Distribution Agreement, or MADA. Google’s proprietary apps — the Play Store, Google Search, Chrome, Gmail, Maps, YouTube — are not part of open-source Android. A manufacturer wanting them had to sign a MADA, and the MADA came as a bundle: to get the Play Store, the commercially indispensable component, the manufacturer had to pre-install the Google Search app and, from 2012, Chrome, and to place them prominently. No Play Store licence was available without them.

The second layer was the Anti-Fragmentation Agreement, or AFA, later reworked as the Android Compatibility Commitment. Any manufacturer wanting to pre-install Google apps on even one device model had to promise not to sell any device, across its whole portfolio, running a version of Android that failed Google’s compatibility tests. In practice this meant that a company like Samsung could not sell Google-approved phones in Europe while also selling a Fire-OS-style Android fork elsewhere in its lineup. The AFA converted Google’s hold over individual devices into a hold over entire companies.

The third layer was the revenue share agreement, or RSA. Google shared a slice of its search advertising revenue with manufacturers and mobile network operators in exchange for exclusive pre-installation of Google Search — initially across a partner’s whole device portfolio, later on a device-by-device basis. Where the MADA guaranteed presence, the RSA purchased absence: absence of any competing search service on the device.

Seen from Google’s side, the model was a rational and even elegant solution to a coordination problem. Free OS, thriving device competition, low handset prices, a consistent app platform for developers, all funded by advertising. Google has argued throughout the case, and repeated after the judgment, that Android expanded choice rather than restricting it, that it created the only serious counterweight to Apple’s iOS, and that the Commission ignored competition from Apple by defining markets narrowly around licensable operating systems and Android app stores.

Seen from the Commission’s side, the same model looked like a moat-building machine. By 2018, Android ran on roughly 80 percent of European smartphones. Google Search held a share above 90 percent in general search in most member states. The contracts did not merely distribute Google’s apps; they made it commercially irrational for any manufacturer to ship a major alternative and contractually impossible to ship a de-Googled fork. The Commission’s phrase for the overall design was a “single and continuous infringement” — three distinct practices serving one strategy, namely cementing the dominance of Google Search during the pivot to mobile.

The courts had to decide which of these two descriptions matched the law’s definition of abuse. Both the General Court in 2022 and the Court of Justice in 2026 concluded, with limited exceptions, that the Commission’s description was the accurate one. The business model was lawful in its generosity — nobody fined Google for giving Android away — and unlawful in its conditions. That distinction, between a free product and the strings attached to it, is the intellectual center of the entire case, and it is the part with the longest reach into how other platform companies structure their ecosystems.

The 2018 Commission decision and its three abuse findings

The formal decision, Case AT.40099 Google Android, adopted on 18 July 2018 under Article 102 of the Treaty on the Functioning of the European Union and Article 54 of the EEA Agreement, runs to hundreds of pages. Its analytical spine is a set of market definitions, a finding of dominance in each defined market, and three separate abuses knitted into one overall strategy.

The Commission defined four relevant markets. First, the worldwide market, excluding China, for licensable smart mobile operating systems — operating systems a manufacturer can licence and install, which excludes Apple’s iOS because Apple licenses it to no one. Second, the worldwide market, excluding China, for Android app stores. Third, the national markets for general search services in each EEA country. Fourth, the worldwide market for non-OS-specific mobile web browsers, relevant to the Chrome analysis. China was carved out because Google’s services barely operate there and the competitive conditions differ fundamentally.

On those definitions, dominance followed almost arithmetically. Android held more than 95 percent of the licensable OS market. The Play Store accounted for more than 90 percent of app downloads on Android devices. Google Search exceeded 90 percent in general search in most member states. Google contested the definitions fiercely — its central objection was always that excluding Apple from the OS market assumed away the most important competitive constraint it faced — but the Commission held that from a manufacturer’s perspective, iOS was simply not an alternative one could licence, and from a user’s perspective, switching between ecosystems was rare and costly.

The first abuse was the tying of Google Search and Chrome to the Play Store through the MADA. In the Commission’s framework, the Play Store was the tying product: no serious Android phone could ship without it, because a phone without the Play Store cannot easily get apps, and a phone that cannot get apps does not sell. Google Search and Chrome were the tied products. Manufacturers could not obtain the tying product without accepting the tied ones. The Commission found this gave Google’s search and browser apps a distribution advantage no rival could match. Rivals could still be installed alongside, in theory; in practice, pre-installation created usage patterns competitors could not overcome by offering downloads. The Commission documented that users overwhelmingly stick with pre-installed defaults, the phenomenon the courts would later discuss as status quo bias.

The second abuse was the anti-fragmentation obligation. The Commission was careful about scope here. It did not challenge Google’s right to require compatibility on devices carrying Google’s own apps; a company may protect the integrity of its platform where its products ship. What it condemned was the extension of that requirement to devices on which no Google app was installed. A manufacturer that wanted Google apps on any device was forbidden from selling incompatible forks on all devices. The Commission found this had concretely blocked fork-based ecosystems from reaching the market — the most cited casualty being Amazon’s Fire OS, which struggled to find major manufacturers willing to build for it because building for Amazon meant losing Google.

The third abuse concerned the revenue share agreements, specifically the portfolio-based RSAs in force between 2011 and 2014, under which payments were conditional on the exclusive pre-installation of Google Search across a partner’s entire device portfolio. The Commission treated these as exclusivity payments in the tradition of the case law on loyalty rebates: payments whose purpose and effect was to ensure no rival search app got shelf space.

Binding it all together was the finding of a single and continuous infringement running from January 2011 to July 2018 — a strategy, executed through complementary contracts, to defend Google Search’s dominance against the threat and the opportunity of mobile. The decision gave Google 90 days to end the conduct and set the fine at €4.34 billion. Google announced its appeal almost immediately, and in October 2018 the case moved to the General Court in Luxembourg, where it would sit for four years. In parallel, Google changed its European contracts: it unbundled Search and Chrome from the Play Store licence, introduced separate paid licensing for its app bundle in the EEA, dropped the condemned exclusivity terms, and later rolled out a search engine choice screen on new Android devices. Those changes meant the market moved on while the litigation stood still — a tension that runs through every assessment of what the case achieved.

The €4.125 billion figure and the arithmetic behind it

The number attached to this case has been reported so often that its internal logic tends to disappear. It is worth reconstructing, because the arithmetic explains both why the fine was a record and why the General Court’s later reduction was modest.

The Commission’s fining methodology under its 2006 guidelines starts from the value of the company’s sales connected to the infringement in the last full year of the violation. For the Android decision, that meant the value of relevant Google sales within the European Economic Area during 2017, essentially the search advertising revenue that flowed through the distribution channels the contested agreements protected. The Commission applied a gravity coefficient of 11 percent to that sales value, multiplied the result by roughly 7.52 to reflect the number of years the infringement lasted, and then added a deterrence amount equal to a further 11 percent of the 2017 sales value. Gravity percentages can run up to 30 percent for the most serious cartels; 11 percent signals a serious but not maximal classification. Duration did most of the heavy lifting. An infringement running from early 2011 into 2018 multiplies whatever the annual base is by more than seven, which is how a percentage-based formula produces a nine-zero outcome.

The original 2018 total was €4,342,865,000. When the General Court partially annulled the decision in 2022, striking down the findings on one category of revenue share agreements, it recalculated the fine itself rather than sending the case back, exercising its unlimited jurisdiction over penalties. The recalculated figure was €4,125,000,000, a reduction of roughly €218 million, or about 5 percent. The court explained the modest size of the cut by noting that the deliberate nature of the remaining infringement, rather than mere negligence, justified keeping the penalty close to its original level even after one pillar of the decision fell. Within the original 2018 total, the Commission had made Alphabet, the parent company, jointly and severally liable alongside Google LLC for €1,921,666,000, reflecting the period after the corporate restructuring that created Alphabet in 2015; when the General Court reset the fine at €4.125 billion, it revised that shared portion to €1,520,605,895, and the Court of Justice confirmed both figures as they stood.

Two comparisons put the sum in perspective. Against the history of EU competition enforcement, it remains the largest antitrust fine ever imposed, ahead of the €2.42 billion Google Shopping penalty of 2017 and the €2.95 billion ad tech fine of September 2025. Against Alphabet’s own accounts, it is small. Alphabet generates more than €4.125 billion in profit in a matter of weeks, and the company recognised the financial impact years ago, meaning the confirmed fine will not visibly dent any future quarterly report. Analysts noted that Alphabet shares slipped around 1 percent in premarket trading after the judgment, a movement indistinguishable from ordinary market noise.

That asymmetry between record-setting nominal size and negligible financial pain is itself one of the case’s lessons, and it is a lesson the European Commission absorbed while the appeals were still running. The Digital Markets Act, which entered into application in 2023, allows fines of up to 10 percent of worldwide turnover for a first infringement and 20 percent for repeat offences, and it front-loads obligations rather than waiting years for effects analysis and litigation. The Android fine’s deterrent force was always going to be reputational and doctrinal rather than financial. Its confirmation now converts eight years of provisional condemnation into permanent legal fact, and permanence is worth more to enforcers than any plausible increase in the amount.

There is one further financial wrinkle worth recording. Because Google paid the fine provisionally after the 2018 decision, as companies typically do to stop interest accruing, the confirmation does not trigger a new payment. Had Google won, the Commission would have owed the money back with interest, a scenario EU courts have grappled with in other cases where annulments produced substantial repayment obligations. The dismissal extinguishes that contingency and closes the account for good.

Android’s free-but-not-free bargain with the phone industry

To understand what the courts condemned, it helps to understand the commercial machine that Android is, because the case was never about the operating system’s source code. The Commission said so explicitly in 2018: the decision did not challenge Android as an open-source project or the licensing of the operating system itself. It challenged the contracts wrapped around it.

Android’s founding bargain, struck when Google bought the startup in 2005 and accelerated after the iPhone’s 2007 debut, was that the operating system would cost phone makers nothing. Any manufacturer can take the Android Open Source Project code, modify it, and ship it without paying Google or even talking to Google. That freedom is real, and it is the foundation of Google’s public defence throughout the case. But an operating system without applications is a commercial shell, and by the early 2010s the applications that mattered to European consumers were overwhelmingly distributed through one channel: Google Play, the app store that manufacturers cannot replicate and consumers will not do without. A phone that cannot run the Play Store is, for mainstream European retail purposes, not a competitive product. Amazon’s Fire phones and tablets demonstrated the point by trying to live without it.

Google’s proprietary layer, the Google Mobile Services suite containing the Play Store, Search, Chrome, Maps, YouTube, Gmail and the underlying play services framework that many third-party apps depend on, was never open source. Access to it required signing Google’s agreements. That is where the power sat. The operating system was free; the ecosystem that made the operating system sellable was conditional. The Commission’s case, in essence, was that Google used the indispensability of the Play Store as a lever to guarantee distribution for two other products, Search and Chrome, which faced genuine competition and did not need to be part of any technical bundle.

The economics behind the bargain are not mysterious, and Google has never hidden them. Android exists to protect and extend Google’s search advertising business in the mobile era. Giving the operating system away made Android the default choice for every manufacturer other than Apple, which drove Android to a share of European smartphone sales exceeding 70 percent through most of the infringement period, and in some national markets above 80 percent. Every one of those devices shipping with Google Search in the default search box, and Chrome as the default browser feeding queries to the same search engine, converted operating system share into search query share, and search query share into advertising revenue. The free operating system was the customer acquisition cost of the world’s most profitable advertising machine.

None of that is illegal. Dominance itself is not an infringement in European law, and neither is building an integrated ecosystem. What Article 102 prohibits is a dominant company using its position in one market to foreclose competition in another through means other than merit. The legal question the case posed was whether the contractual conditions attached to the free bargain crossed that line. The Commission said they did, in three distinct ways that reinforced each other. The General Court agreed on two of the three. The Court of Justice has now made that assessment final. The next three sections take each mechanism in turn, because the differences between them explain both the outcome and its limits.

Pre-installation conditions and the mechanics of the MADA contracts

The Mobile Application Distribution Agreement deserves close attention, because it was the load-bearing wall of Google’s mobile strategy and the primary target of the tying finding. Understanding its mechanics explains why the Commission saw coercion where Google saw a standard licensing bundle.

A MADA was, on its face, a free licence. Google charged manufacturers nothing for the suite of Google Mobile Services. In exchange, the manufacturer accepted conditions on what had to be installed and where it had to appear. The core obligations, as documented in the Commission decision and the court judgments, were that a device licensing the Play Store must also pre-install the Google Search app and the Chrome browser, that the Search widget be given prominent default placement, and that Google Search be set as the default search service in specified access points. The agreements covered device models individually, but in commercial reality no major manufacturer could pick and choose: a flagship phone without the Play Store was unsellable in Europe, so the conditions applied to effectively the entire Android device market.

Two features of this arrangement mattered enormously to the legal analysis. The first is that the bundle was non-negotiable in the only direction that counted. A manufacturer could add rival apps alongside Google’s — Samsung shipped its own browser and later Bing deals were floated at various points — but it could never subtract Google’s. The floor of Google’s distribution was guaranteed on essentially every Android phone sold in the EEA. Rivals started every device generation from zero while Google started from one hundred percent. In markets where user behavior is governed by defaults, a guaranteed floor of presence converts into a near-guaranteed ceiling for competitors.

The second feature is the economic asymmetry between pre-installation and download. Google argued throughout the case that any user could download a competing search app or browser in seconds, free of charge, and that this made the pre-installation advantage trivial. The Commission answered with usage data: on Android devices, where Google Search and Chrome were pre-installed, Google’s search share was materially higher than on Windows Mobile devices of the era, where Bing was the pre-set default and Google had to be sought out. Very few users on either platform changed their defaults. The evidence showed that defaults, not quality comparisons, decided the outcome for the overwhelming majority of users — a finding that would echo years later, almost verbatim, in the United States government’s case against Google’s search distribution payments to Apple.

The General Court engaged with this evidence at length in 2022 and accepted the Commission’s core proposition: pre-installation conferred a substantial competitive advantage that could not be offset by downloads, and Google’s own documents showed the company understood this perfectly well when designing the contracts. Internal Google material cited in the proceedings treated default placement as decisive commercial territory, which is difficult to reconcile with a litigation position that placement barely matters.

Before the Court of Justice, Google reframed the attack as a matter of legal method rather than fact. It argued the General Court had wrongly assessed the MADA pre-installation conditions in combination with the revenue share agreements when evaluating effects, and that the Commission had been obliged to conduct a counterfactual analysis — a rigorous comparison with a hypothetical world without the conduct — before it could find abuse. The Court of Justice rejected both limbs. It held that the General Court was entitled to take account of the relevant economic context in its entirety, including the revenue share agreements, and that establishing an abuse of dominance does not systematically require a counterfactual exercise. Conduct, context and capability to foreclose can suffice. That methodological ruling is one of the judgment’s most consequential exports to future cases, because counterfactual requirements are the standard defensive weapon of dominant firms in effects-based litigation.

For the industry, the MADA finding drew a durable line. Bundling a must-have platform component with adjacent products the platform owner wants distributed is now, for a dominant firm in Europe, presumptively dangerous territory. Microsoft learned a version of this lesson with Windows Media Player and Internet Explorer two decades ago; the Android case translated it into the mobile era and confirmed that the doctrine survived the shift intact.

Anti-fragmentation agreements and the fate of Android forks

The second abuse is the least understood by the general public and arguably the most strategically important. The anti-fragmentation agreements did not put Google’s apps anywhere. They kept other people’s operating systems off the market.

Android’s open-source licence permits anyone to take the code and modify it. A modified, independently maintained version is a fork. Forks are the built-in safety valve of open source: if the steward of a project turns extractive, the community can route around it. In Android’s case, forks were also the most credible path for a rival ecosystem to emerge without the impossible task of writing a mobile OS from scratch. Amazon proved the concept with Fire OS, a fully functional Android fork with its own app store, its own services layer, and no Google inside.

The AFAs neutralized this safety valve. Under them, any manufacturer that pre-installed Google apps on any device committed not to manufacture or sell any device running an incompatible fork. Compatibility was defined by Google’s own test suite. The commitment was portfolio-wide and, given that every major manufacturer needed Google apps on its mainstream phones, it bound essentially the entire manufacturing base capable of producing quality hardware at scale. When Amazon sought manufacturing partners for Fire OS phones and tablets, the large Android manufacturers could not build for it without forfeiting their Google licences across their whole business. The Commission documented this mechanism concretely, and the courts accepted that the AFAs had hindered the development of forks that could have become distribution channels for rival search services.

Google’s defense rested on fragmentation itself as a genuine engineering problem. A platform splintered into incompatible variants harms developers, who must test against endless permutations, and users, whose apps break. Google pointed to the cautionary tale of early Unix and to Symbian’s chaos. The compatibility program, it argued, was what made Android a coherent platform worth developing for, and the whole ecosystem — including the Commission’s imagined fork-based competitors — benefited from it.

The Commission’s response was surgical rather than dismissive, and it is worth restating because commentary often gets it wrong. The decision did not condemn compatibility requirements as such. Google remained free to insist that any device carrying Google apps pass its compatibility tests. What the decision condemned was the extension of the restriction to devices on which no Google app was installed. A fork device without Google Search, without the Play Store, without Chrome, poses no integrity risk to Google’s products, because Google’s products are not on it. The only thing such a device threatens is Google’s market position. The General Court agreed that objective justification failed for precisely this reason, and the Court of Justice confirmed in 2026 that the pre-installation and anti-fragmentation agreements restricted competition within the Android ecosystem and strengthened Google’s dominant position, finding no error in the lower court’s analysis.

The real-world consequences of the AFA finding surfaced gradually. After 2018, Google reworked the arrangements in Europe, and manufacturers gained formal freedom to build forked devices. The results have been modest — no European fork ecosystem materialized — which both sides read as vindication. Google says the market spoke: nobody wanted forks. Critics say the window closed: by the time the freedom arrived, network effects around the Play Store were insurmountable, and the harm the AFAs did in 2011 through 2018, when a fork ecosystem was still viable, could no longer be undone. The most visible fork success since, Huawei’s HarmonyOS, was born of US sanctions rather than European remedies, and its European market presence remains small. The episode stands as the case’s clearest illustration of a hard truth about antitrust enforcement in fast markets: a remedy that arrives after the critical window has closed restores the law without restoring the competition.

Revenue share agreements and the part the courts trimmed away

Google did not lose everything across eight years of litigation, and the exception matters for understanding what the final judgment does and does not establish. The exception concerns the revenue share agreements, the third pillar of the Commission’s decision.

Between 2011 and 2014, Google operated portfolio-based RSAs with certain manufacturers and mobile network operators. Under them, Google paid a percentage of the search advertising revenue generated on the partner’s devices, on condition that Google Search was the only general search service pre-installed across the partner’s entire relevant portfolio. From 2014 Google moved to device-based agreements, which conditioned payment on exclusivity for particular devices rather than whole portfolios; the Commission’s abuse finding targeted the portfolio-based generation.

The Commission analyzed these payments within the tradition of exclusivity rebates: money paid by a dominant firm not for performance but for the foreclosure of rivals. On that view, the RSAs bribed the distribution channel shut. The going rate for placing a rival search app on a major manufacturer’s home screen was not merely the placement fee a rival would pay; it was that fee plus the entirety of the Google revenue share the partner would forfeit. The Commission concluded that payments structured this way were capable of making it uneconomical for rivals as efficient as Google to compete for placement.

The General Court disagreed — not with the theory, but with the execution. Applying the Court of Justice’s 2017 Intel judgment, which requires the Commission to seriously examine a dominant firm’s evidence that its rebates could not foreclose an as-efficient competitor, the General Court found the Commission’s economic analysis of the RSAs defective. Among other flaws, the court held that the Commission had erred in its assessment of the costs a hypothetical rival would bear and of the share of the search market the agreements actually covered, and had not properly established that the payments were capable of foreclosing an as-efficient competitor. The General Court annulled the RSA abuse finding in 2022, and this annulment is the reason the fine fell from €4.343 billion to €4.125 billion. The Commission did not cross-appeal, so the annulment stood untouched before the Court of Justice, and it survives in the final outcome.

The surviving annulment has been seized on by both camps. Google’s sympathizers present it as proof that the Commission overreached and that at least one third of its theory collapsed under scrutiny. Enforcement advocates note the narrowness of the loss: the court faulted the quality of one economic analysis, not the principle that exclusivity payments by dominant firms can be abusive, and the two structural abuses — tying and anti-fragmentation — emerged fully intact and are now confirmed at the highest level. The fine dropped by five percent; the strategy stood condemned.

For practitioners, the RSA chapter carries a distinct lesson about evidentiary standards. Where the Commission builds a case on the architecture of contracts — what they require, what they forbid, what they make commercially impossible — the courts have proven willing to sustain it on qualitative evidence and economic context. Where the Commission builds a case on price-cost mathematics, as the as-efficient-competitor framework demands for payment-based conduct, the calculations must withstand forensic attack, and in the Android RSAs they did not. That asymmetry now shapes how Brussels drafts decisions: structural theories where possible, quantitative ones only where the numbers are bulletproof. The Court of Justice’s later handling of the as-efficient-competitor question in this very case, discussed below, pushed the doctrine further in the Commission’s favor.

The 2022 General Court judgment and the reduced fine

The General Court’s judgment of 14 September 2022 in case T-604/18 was the pivotal middle act of the litigation, and its structure explains everything that happened afterward. The court sat in an expanded five-judge formation, a signal of the case’s importance, heard five days of oral argument in the autumn of 2021, and delivered a judgment of more than a thousand paragraphs.

On the market definitions, the court sided with the Commission across the board. It accepted the licensable-OS market that excluded Apple, accepting the Commission’s evidence that iOS exerted only indirect and insufficient competitive pressure on Google’s conduct toward manufacturers. It accepted the Android app store market and the national search markets. Google’s most fundamental line of attack — that the whole case rested on gerrymandered markets designed to make Google look dominant — failed at the first judicial level and was never seriously revived.

On the tying abuse, the court upheld the Commission in full. It confirmed that the Play Store and the Search app were distinct products, that the MADA conditions constituted tying, and that the tying was capable of restricting competition given the significance of pre-installation, the reality of status quo bias among users, and the inability of rivals to offset the advantage through downloads or agreements of their own. It likewise upheld the Chrome tying finding and the anti-fragmentation abuse, including the rejection of Google’s objective justifications.

On the RSAs, as described above, the court annulled. And on the fine, it did something subtle: rather than simply reducing the amount proportionally to the annulled part, the court exercised its unlimited jurisdiction over penalties, made its own assessment of the gravity and duration of the infringement it had upheld, and fixed the fine at €4.125 billion — a figure that reflected the court’s slightly different reasoning while remaining within touching distance of the original. The message embedded in that arithmetic was pointed: even shorn of the RSA finding, the conduct the court had confirmed justified a penalty of historic size.

The 2022 judgment mattered beyond its operative part because of how it handled evidence. The court engaged deeply with the empirical record — usage statistics, internal Google documents, the Windows Mobile comparison, testimony about manufacturer incentives — and found the Commission’s factual case solid. Appeals to the Court of Justice lie on points of law only; findings of fact made by the General Court are, with narrow exceptions for distortion of evidence, final. By losing the factual battle comprehensively in 2022, Google entered the final appeal with its hands tied, able to argue only that the General Court had applied wrong legal tests to facts it could no longer contest.

Google’s public reaction at the time previewed the arguments it would carry to Luxembourg’s higher floor. The company expressed disappointment, repeated that Android had created more choice for everyone, and filed its appeal to the Court of Justice in December 2022, opening case C-738/22 P. The Commission, for its part, chose not to appeal the RSA annulment — a decision some competition lawyers criticized as leaving doctrinal ground undefended, but which in hindsight looks shrewd: it narrowed the final round to terrain where the Commission’s position was strongest, and the Court of Justice’s eventual reasoning on the as-efficient-competitor principle recovered much of what the RSA annulment had conceded.

Google’s six grounds of appeal before the Court of Justice

An appeal to the Court of Justice is not a retrial. It is an attack on legal reasoning, and Google’s lawyers constructed theirs on six grounds, each aimed at a joint in the General Court’s analysis. The grounds reward examination because the court’s answers to them are the doctrine the case now stands for.

The first ground targeted the effects analysis of the MADA pre-installation conditions. Google argued the General Court had erred by assessing those conditions together with the revenue share agreements — including the portfolio RSAs whose abuse finding had been annulled — when evaluating anticompetitive effects. If the RSA finding fell, Google reasoned, any effects analysis that leaned on the RSAs as context had to fall with it. The argument had a seductive logic: how can annulled conduct aggravate surviving conduct?

The second ground pressed the as-efficient-competitor point. Google contended that the General Court wrongly endorsed findings that the MADAs could exclude competitors as efficient as Google without the Commission having properly demonstrated it, importing into the tying analysis the quantitative rigor the Intel line of cases demands for pricing conduct.

The third and fourth grounds attacked the treatment of the anti-fragmentation agreements, challenging both the finding that restricting incompatible forks restricted competition — Google maintained the forks kept off the market were not shown to be genuine competitive threats — and the rejection of its objective justification rooted in protecting the ecosystem from fragmentation.

The fifth and sixth grounds went to the fine: the methodology of the General Court’s recalculation and the proportionality of a penalty barely reduced despite a partly annulled decision.

The Court of Justice dismissed all of them. On the first ground, it held that the General Court was entitled to take account of the relevant economic context in its entirety, including the revenue share agreements, when assessing the effects of the pre-installation conditions. Annulment of the RSA abuse finding for deficient economic analysis did not erase the RSAs from commercial reality; they remained part of the environment in which the tying operated, and a court assessing effects may look at the world as it was. It further confirmed that no systematic counterfactual analysis is required to establish an infringement of Article 102 — a holding that removes a demand dominant firms have pressed in nearly every effects-based case since Intel.

On the second ground, the court drew a boundary around the as-efficient-competitor framework. It confirmed the General Court’s approach that where conduct like tying is concerned, and where the structure of the market means no real competitor could match the dominant firm’s advantages — Google’s ubiquity of pre-installation, its network effects, its data — insisting on a hypothetical as-efficient-competitor comparison would be artificial. The court found there was a status quo bias in favor of pre-installed apps and that Google had not demonstrated that user preferences or app quality, rather than the contractual architecture, explained its position. On the AFA grounds, it confirmed both the restriction finding and the failure of the justification. On the fine, it found the General Court’s exercise of its unlimited jurisdiction unimpeachable.

Nothing in the operative part gave Google even partial relief. In the vocabulary of appellate litigation, it was a whitewash — six grounds raised, six grounds dismissed, costs awarded against the appellants.

Advocate General Kokott’s opinion and the as-efficient-competitor question

Between the General Court’s judgment and the final ruling stood a document that told careful observers how the story would end. On 19 June 2025, Advocate General Juliane Kokott delivered her opinion in C-738/22 P, recommending that the Court of Justice dismiss Google’s appeal in its entirety. Advocate General opinions do not bind the court, but the court follows them in the substantial majority of cases, and Kokott — one of the longest-serving and most influential advocates general in the institution’s history, with a track record of muscular readings of Article 102 — left little room for suspense.

The intellectual heart of her opinion was the as-efficient-competitor question, and it deserves unpacking because it is where the Android case reshapes European competition doctrine most visibly. The as-efficient-competitor principle, in its classic form, holds that competition law protects competition, not competitors: a dominant firm does not abuse its position merely by outperforming weaker rivals, and conduct should generally be condemned only if it could exclude a rival as efficient as the dominant firm itself. Dominant firms have wielded the principle as a shield, arguing that unless the authority proves a hypothetical equally efficient rival would be foreclosed, no abuse exists.

Kokott’s opinion confronted the fiction at the center of that test when applied to a company like Google. She concluded that “it is not realistic to compare the situation of Google with that of a hypothetical as-efficient competitor”, because Google’s position rests on dominance across multiple interlocking digital markets and on network effects that no hypothetical entrant could replicate. An as-efficient competitor to Google is not merely hypothetical; it is structurally impossible, and demanding proof of its foreclosure would immunize precisely the firms whose conduct matters most. Insisting on such comparisons in digital markets, she argued, risks undermining the effectiveness of EU competition law altogether.

The Court of Justice absorbed the substance of this reasoning. Its judgment confirms that the as-efficient-competitor criterion is one analytical tool among several, not a mandatory gateway to every abuse finding, and that in markets shaped by network effects, data advantages and default-driven user behavior, an authority may establish abuse through the capability of the conduct to foreclose, assessed in its full economic context. For future defendants, the practical meaning is stark: the strongest firms derive the least protection from the as-efficient-competitor shield, because their very strength makes the comparison unrealistic.

Kokott’s opinion also mattered institutionally. It arrived at a moment when transatlantic tension over technology regulation was rising, with Washington periodically characterizing EU enforcement against American firms as disguised industrial policy. An opinion of that rigor from the court’s own advocate general, methodically dismantling each of Google’s six grounds on doctrinal terms, armored the eventual judgment against the charge of politics. When the Second Chamber delivered its ruling on 2 July 2026, tracking her recommendation, the outcome could be presented — accurately — as the consistent product of a decade of case law rather than a verdict improvised against a convenient foreign target.

Inside the 2 July 2026 judgment

The judgment itself, delivered by the Second Chamber of the Court of Justice, is a dense appellate text, but its load-bearing holdings can be stated with precision, and several of them will be quoted in competition filings for decades.

The first holding concerns context. The court ruled that the General Court did not err in law when assessing the anticompetitive effects of the pre-installation conditions laid down by the Android agreements, because a court may take account of the relevant economic context in its entirety — including agreements, like the RSAs, whose separate condemnation had been annulled. Abuse analysis under Article 102 is not conducted contract by contract in hermetic isolation; a strategy is judged as a strategy. This holding rescues the Commission’s single-and-continuous-infringement framing from the fragmentation Google’s first ground attempted, and it licenses future decisions to describe ecosystems of mutually reinforcing agreements rather than itemized clauses.

The second holding concerns method. The court confirmed that establishing an infringement of the prohibition on abuse of dominance does not require the systematic performance of a counterfactual analysis. The Commission must show that conduct is capable of restricting competition on the merits, using evidence appropriate to the conduct and the market; it need not construct and defend a full model of the world that would have existed absent the conduct. Dominant firms had pressed the counterfactual requirement in case after case since Intel, and the Android judgment closes that door for conduct of this type.

The third holding concerns users. The court confirmed the General Court’s finding that a status quo bias operates in favor of pre-installed apps — users keep what ships on the device — and, critically, that Google had not demonstrated that user preferences or the alleged superior quality of its apps, rather than the pre-installation architecture, explained its distribution advantage. The burden allocation embedded in that sentence matters: once the authority shows the conduct’s capability to foreclose through defaults, it falls to the dominant firm to prove that merit, not mechanics, drove the outcome. Google, with the deepest behavioral dataset in the industry, could not carry that burden, which is itself a finding with evidentiary poetry to it.

The fourth holding concerns the anti-fragmentation agreements. The court confirmed that the General Court was entitled to conclude the AFAs restricted competition within the Android ecosystem and strengthened Google’s dominant position, and that Google’s ecosystem-integrity justification failed insofar as the restriction reached devices carrying no Google apps at all. Platform stewardship, in other words, justifies rules that protect the steward’s products where they are present; it does not justify rules whose only function is to prevent rivals from existing.

The fifth holding concerns the penalty. The court found no error in the General Court’s recalculation of the fine under its unlimited jurisdiction, confirming both the €4.125 billion figure and the methodology behind it. The proportionality attack — the claim that losing a third of the decision should cost the Commission more than five percent of the fine — failed, cementing the principle that the fine attaches to the gravity of what is upheld, not to a scorecard of annulled paragraphs.

The operative result: appeal dismissed in full, penalty confirmed as revised by the General Court, Google and Alphabet to bear the costs. The court’s press communication compressed eight years into a single sentence — the appeal is dismissed, “thereby confirming the penalty imposed for Google Search’s abuse of a dominant position in the context of the Android operating system.”

Reading the judgment as a whole, its temperament is as telling as its holdings. There is no rhetorical flourish, no digital-age throat-clearing, no gesture toward the political weight of the defendant. The court treats Google’s contracts the way it treated Hoffmann-La Roche’s loyalty rebates in 1979 and Microsoft’s bundles in the 2000s: as commercial instruments to be measured against a stable prohibition. That continuity is the quiet message to every platform general counsel in the world. European abuse-of-dominance law did not need to be reinvented for the smartphone era; it needed only to be applied, and the institutions proved willing to apply it against the largest advertising business ever built.

Status quo bias and the economics of default settings

Strip away the case numbers and the Latin, and the Android litigation was fundamentally an argument about a single behavioral fact: people do not change defaults. The entire multi-billion-euro edifice — the contracts, the fine, the eight years of appeals — rests on the question of whether pre-installed placement determines market outcomes or merely decorates them. The courts answered that it determines them, and the evidentiary basis for that answer is worth understanding, because it now functions as judicially established fact in European law.

Behavioral economics has documented default effects across every domain where they have been measured. Organ donation rates diverge by tens of percentage points between opt-in and opt-out countries with similar cultures. Retirement savings participation jumps when enrollment is automatic. Software settings persist untouched through years of use. The mechanism is not stupidity but rational economizing: evaluating alternatives costs time and attention, the default carries an implicit endorsement, and for most choices the stakes feel too low to justify the search. A search engine default is a textbook case — the pre-set option is adequate, the alternative is unfamiliar, and the cost of switching, while objectively tiny, exceeds the perceived benefit for almost everyone.

The Commission did not rely on the literature alone. It assembled market-specific evidence, and the most compelling piece was the natural experiment that Windows Mobile provided. On Android devices, where Google Search shipped as the default, Google’s mobile search share was overwhelming. On Windows Mobile devices of the same era, where Microsoft’s Bing was pre-set, Google’s share of searches collapsed to a fraction of its Android level — while Bing, a product with negligible share anywhere users chose freely, captured the large majority of queries. Same users, same needs, same Google available one download away; radically different outcomes, driven by nothing but the factory setting. Few datasets in competition history have isolated a causal mechanism so cleanly.

Google’s counterargument was that its search share was equally overwhelming on iOS, where the Commission’s condemned contracts did not apply — proof, it said, that quality rather than contracts explained everything. The response, accepted by the courts, was that Google’s iOS position was itself purchased through default placement, via the multibillion-dollar annual payments to Apple that would later anchor the US Department of Justice’s case. Everywhere one looked, Google’s mobile dominance traveled with default status; nowhere did it demonstrably survive without it. The Court of Justice’s confirmation that Google failed to prove user preference or quality as the true explanation is the final judicial word on that dispute.

The doctrinal residue is the phrase now embedded in the case law: status quo bias. Its recognition converts a behavioral regularity into a legal presumption with teeth. When a dominant firm secures defaults through contract, European authorities may treat the foreclosure as real without waiting for body counts among rivals, and the firm bears the burden of proving that merit would have produced the same result. Every subsequent fight over choice screens, self-preferencing and pre-installation — under the Digital Markets Act and beyond — inherits that presumption. The economics of attention, long understood in Mountain View better than anywhere on earth, is now understood in Luxembourg too.

Article 102 TFEU and the legal doctrine the ruling cements

The legal foundation of the entire case is a single treaty provision of remarkable brevity. Article 102 of the Treaty on the Functioning of the European Union prohibits, in its operative words, any abuse by one or more undertakings of a dominant position within the internal market insofar as it may affect trade between member states. Everything else — the market definitions, the tying tests, the effects analysis — is judicial construction built on that sentence over sixty years, and the Android judgment is now a permanent part of the structure.

Three doctrinal contributions stand out. The first is the consolidation of tying law for the platform age. The classic four-part test — distinct products, dominance in the tying market, coercion, and foreclosure effects — derives from cases about vitamins, cash registers and media players. The Android judgment applies it to a configuration those cases never imagined: a tying product given away free, a tied product also free, and harm measured not in overcharge but in the diversion of attention and data. The courts held that the absence of a monetary price changes nothing in the analysis when the products are paid for in distribution and defaults. For an economy in which the most powerful products are nominally free, that is the single most important sentence in the case.

The second contribution is the recalibration of the as-efficient-competitor principle described earlier. After Intel in 2017, dominant firms read the case law as installing a quantitative gate in front of every abuse finding. The Android judgment, following Advocate General Kokott, confines that reading: the criterion applies naturally to pricing conduct where cost comparisons make sense, but where dominance rests on network effects, data scale and default architecture that no equally efficient rival could possess, the comparison is unrealistic and cannot be demanded. Abuse may be established through capability to foreclose in context. Whatever one’s view of the merits, the direction is unmistakable — the evidentiary path to condemning ecosystem conduct by super-dominant platforms is shorter after Android than before it.

The third contribution concerns objective justification, the escape hatch through which dominant firms may defend restrictive conduct as necessary and proportionate to a legitimate aim. The AFA analysis gives the defense a precise shape: legitimate platform-integrity interests justify restrictions that protect the platform where the firm’s products operate, and stop justifying them at the exact point where the restriction’s only remaining function is the suppression of competitors. Proportionality is measured clause by clause, device by device. General counsels drafting ecosystem rules now have a judicial template for where lawful stewardship ends.

Beneath the doctrine sits a philosophical settlement worth naming. For two decades, an argument ran through competition policy that digital markets were different — too fast, too fluid, too innovation-driven for structural antitrust, self-correcting through disruption in ways that made intervention more dangerous than dominance. The European institutions never fully accepted the argument, and the Android case is its formal burial in EU law. The courts examined the fastest-moving consumer market in history and found ordinary things: contracts that excluded, payments that foreclosed, defaults that decided. The tools of 1979 proved serviceable in 2026. Jurisdictions from Brazil to India to Japan that model their competition statutes on Article 102 will import that settlement along with the holdings, which is how a Luxembourg judgment about European phone contracts becomes a global reference point.

Margrethe Vestager’s legacy and the trilogy of Google fines

No account of the Android case is honest without its protagonist. Margrethe Vestager arrived as European Commissioner for Competition in November 2014, inheriting a Google investigation — the Shopping case — that her predecessor had tried three times to settle. She abandoned settlement, issued charges, and over the following five years built the largest enforcement program ever directed at a single company: three decisions, three record or near-record fines, and a public profile no competition official had held before her.

The trilogy had a deliberate architecture. Shopping, decided in June 2017 with a €2.42 billion fine, condemned self-preferencing — Google’s demotion of rival comparison-shopping services in its search results while promoting its own. Android, decided in July 2018 at €4.34 billion, condemned the contractual capture of mobile distribution. AdSense, decided in March 2019 at €1.49 billion, condemned exclusivity clauses in Google’s search-advertising intermediation contracts with large publishers. Read together, the three decisions map the full circuit of Google’s business: how results are ranked, how queries are gathered, how the advertising money flows. Vestager was not sniping at practices; she was auditing a model.

Her critics were loud and bipartisan across the Atlantic. American commentators, and at moments the US government itself, accused Brussels of taxing American success under antitrust cover. A different school of critics attacked from the opposite side, noting that fines representing single-digit percentages of annual profit changed no incentives, that remedies arrived years after the harm, and that Google’s market shares survived every decision essentially intact. Vestager’s own answer to the second critique was legislative: the experience of litigating against Google at the speed of courts convinced her, and the Commission behind her, that ex-post enforcement could not police gatekeepers alone. The Digital Markets Act, which she co-championed and which entered full application in 2024, is the institutional lesson of the Google trilogy written into statute.

The judicial fate of the trilogy, settled progressively after her 2024 departure, has been kind to her. The Court of Justice confirmed the Shopping decision in full in September 2024, cementing self-preferencing as an abuse. Android is now confirmed at €4.125 billion, its two structural pillars intact. Only AdSense stumbled: the General Court annulled that decision in September 2024 on grounds relating to the Commission’s assessment of the contractual clauses, a reminder that the courts audited her work rather than rubber-stamping it — which, paradoxically, strengthens the confirmed cases, since no one can claim the judges were captured.

History will likely remember the fines less than the shift in presumption she engineered. Before Vestager, the operating assumption in most capitals was that platform giants were innovation engines to be admired and occasionally consulted. After her, the assumption in Brussels — increasingly exported elsewhere — is that they are infrastructure owners to be supervised. The 2 July 2026 judgment is the highest court in Europe countersigning that shift on the largest case she ever brought.

Google’s response and the argument the court rejected

Google’s public reaction to the judgment was disciplined, brief and, to anyone who has followed the case, familiar. A spokesperson said the company was disappointed, and repeated the two sentences that have anchored its position since 2018: “Android provides more choice for everyone and supports thousands of businesses,” and the judgment “fails to recognize our significant investment to ensure Android remains open, interoperable and free.” The company added that it had adapted its agreements to comply with the original decision back in 2018 and remained focused on innovation and openness for users, partners and developers. Google also stressed that the case rests on market conditions that no longer exist — a mobile ecosystem of 2011 to 2018 judged in a world of 2026.

The choice argument deserves a fair hearing, because it is not frivolous. Android genuinely did break a duopoly trajectory. Without a free, licensable operating system, the smartphone market of the 2010s plausibly consolidates around Apple at the premium tier and a fragmented, weaker field below it. Android enabled hundreds of manufacturers, drove handset prices down to levels that put smartphones in billions of hands, and gave developers a single target reaching most of humanity. Every element of that story is true, and the Commission never disputed it.

The judgment’s answer is that the story is true and beside the point. The abuse finding does not condemn Android’s existence, its openness or its price; it condemns specific conditions attached to specific contracts — conditions whose function was to ensure that the choice Android created for manufacturers and users never extended to the one dimension that funded Google: search. A manufacturer could choose any hardware, any skin, any price point, and could not choose to ship the Play Store without Google Search. A user could choose among a thousand devices and would meet the same default on every one. Choice flourished everywhere except where it threatened the revenue model. The courts found that pattern was not an accident of the model but its design.

The interoperability-and-investment argument fared no better, for a structural reason embedded in the AFA analysis: the court accepted that protecting a platform’s coherence is legitimate, and found that the condemned clauses reached beyond any coherence interest — to devices carrying no Google software at all. An investment-based justification cannot cover restrictions that protect nothing except market share.

The past-market-conditions argument, finally, misunderstands what the proceeding was. The judgment adjudicates conduct between 2011 and 2018 because that is when the conduct occurred; fines are retrospective by nature. Google is right that the 2026 mobile world — AI assistants, choice screens, the Digital Markets Act — differs from the one the Commission investigated. But that observation cuts against Google as easily as for it: several of the differences exist precisely because the 2018 decision forced contractual change. A defendant cannot comply its way out of liability for the years before compliance.

One further note on posture. Google fought this case for eight years without ever offering commitments to settle it, a contrast with its behavior in the first Shopping investigation. That was a rational bet in 2018 — the doctrine was genuinely unsettled, and the RSA annulment shows judicial review had teeth. The bet has now been called, and the price of making it was not just the fine but eight additional years of a confirmed-infringer narrative, discovery material in the public record, and a final judgment stronger for the Commission than a settlement would ever have been.

Reactions from complainants, rivals and consumer groups

Judgments of this size land in an ecosystem of interested parties, and the reactions on 2 July 2026 traced the case’s original battle lines with precision.

FairSearch, the industry coalition whose 2013 complaint formally triggered the Android investigation, greeted the ruling as vindication of a thirteen-year effort. The group — whose membership over the years included Oracle, Nokia and Microsoft, companies with their own scars from Google — had argued from the beginning that Android’s free price concealed a paid structure: manufacturers paid in obedience what they did not pay in fees. The final judgment adopts that framing in all but vocabulary, and FairSearch’s lawyers can claim, accurately, that a trade-association complaint reshaped global platform law.

European consumer organizations, with BEUC prominent among them, welcomed the confirmation while pressing their standing critique: fines confirmed in 2026 do not compensate users harmed from 2011, and consumer redress in EU competition cases remains slow and rare. Their statements pivoted quickly from the judgment to the Digital Markets Act, urging the Commission to enforce Android-related gatekeeper obligations with the confidence the ruling supplies. For the consumer lobby, the case’s value is less the money than the judicial confirmation that default architecture harms consumers — a finding they now deploy in every DMA consultation.

Rival search and browser companies reacted with a mixture of satisfaction and weariness. DuckDuckGo, Ecosia, Qwant and other independent search providers have spent years arguing that choice screens as implemented after 2018 were designed grudgingly — initially auction-based, buried in setup flows, easily dismissed — and that market shares barely moved as a result. Their consistent position, restated after the judgment, is that the ruling proves the diagnosis was right while the remedy arrived too late and too weak, and that real correction now depends on DMA enforcement, including the 2026 orders on AI assistant access and search data sharing. Browser makers in the Open Web Advocacy orbit made the parallel point about Chrome: the tying finding is final, and Chrome’s mobile dominance is its living monument.

The competition bar and academic commentators, meanwhile, converged on the as-efficient-competitor holdings as the judgment’s exportable core. Law firm client alerts published within hours flagged the same practical warnings: ecosystem contracts will be read as strategies, counterfactual demands will no longer stall effects cases, and a final infringement finding of this breadth is a foundation for private damages litigation across all 27 member states. Several noted the Swedish PriceRunner award of days earlier as the shape of things to come.

From member state governments, reaction was quiet and content. Digital sovereignty currents in Paris, Berlin and Brussels absorbed the ruling as further proof that European institutions can discipline American platforms through law rather than politics — a talking point of particular value in a period of transatlantic friction over technology policy. Washington’s response remained muted, complicated by an inconvenient fact: the US government’s own courts had, in the DOJ search case, found Google’s distribution defaults unlawful under American law, using evidence and logic that rhyme with Luxembourg’s.

The money question and where a €4.1 billion fine actually goes

A recurring public question after every EU mega-fine deserves a concrete answer: who receives the €4.125 billion, and what happens to it?

The destination is the general budget of the European Union. Competition fines are paid to the Commission and booked as other revenue in the EU budget, where they reduce, euro for euro, the contributions member states would otherwise make based on gross national income. The fine does not go to victims, to rival companies, or to a dedicated enforcement fund; it effectively becomes a rebate to national treasuries. On a fine of this size, the arithmetic is tangible — a country contributing a few percent of the EU budget sees its bill fall by a corresponding share of €4.125 billion, real money for a finance ministry even if invisible to citizens.

The mechanics of payment explain why 2 July 2026 produced no dramatic cash movement. Companies fined by the Commission must pay provisionally within three months or provide a bank guarantee covering the amount plus interest while appeals run. Google paid provisionally years ago; the funds have been held by the Commission, invested conservatively, pending the litigation’s outcome. Had Google won, the money would have returned with interest — as happened for Intel after its 2022 annulment. Having lost, Google simply loses its claim to restitution, and the provisional payment becomes definitive. Alphabet’s accounts had provisioned for this outcome long ago, which is why the earnings impact in 2026 is negligible and the share price barely registered the ruling.

For Alphabet’s investors, the relevant financial questions were never about the fine itself. They concern the flows the condemned conduct protected and the exposures the judgment opens. On the first: European search advertising generates tens of billions of dollars annually for Google, and the contracts at issue defended that stream through the decisive platform transition of the century; against that backdrop, €4.125 billion spread over eight years of litigation resembles an insurance premium more than a punishment, a point EU legislators internalized when they attached DMA penalties of up to ten percent of worldwide turnover — and twenty percent for repeat infringements — to gatekeeper obligations. On the second: the judgment’s principal financial legacy is not the fine but the liability finding beneath it, which private claimants can now use without relitigating abuse. That exposure, unlike the fine, is uncapped, unprovisioned in any precise way, and only beginning to take shape.

Follow-on damages and the PriceRunner precedent

The most consequential sentence in EU competition procedure for the years ahead may be a technical one: under the damages directive and settled case law, a final infringement decision of the Commission constitutes binding proof of the infringement before the national courts of every member state. Claimants suing for harm caused by conduct the Commission condemned do not need to prove the abuse; they need to prove their damage and its causal link to it. On 2 July 2026, the Android findings crossed the line from contested to binding, and an entire category of litigation became viable at a stroke.

The timing could not have been more instructive, because days before the CJEU ruling, a Swedish court ordered Google to pay approximately $1.5 billion in damages to PriceRunner, the comparison-shopping service now owned by Klarna, in follow-on litigation flowing from the 2017 Shopping decision. PriceRunner had sued in 2022, immediately after the Shopping infringement finding became final, arguing that Google’s demotion of rival comparison services had suppressed its traffic and revenues for years. The award — among the largest competition damages judgments in European history — demonstrated three things at once: national courts will grant enormous sums on the back of confirmed EU decisions, the damages phase can exceed the original fine’s deterrent effect for individual victims, and the queue forms fast. Google has said it will appeal the Swedish ruling, but the signal to the plaintiff bar was unmistakable.

Mapping the Android findings onto potential claimants produces a long list. Rival general search services — Microsoft’s Bing above all, but also DuckDuckGo, Ecosia, Qwant and Seznam in its home market — can argue the tying and default architecture deprived them of mobile query share during the industry’s formative years, with damages models built on the very Windows Mobile evidence the courts credited. Browser developers, Opera and Mozilla among the plausible names, can make the parallel argument about Chrome’s guaranteed distribution. App store operators and would-be operators can point to the AFA findings; Amazon’s difficulty finding manufacturing partners for Fire OS devices is documented in the decision itself, effectively a pre-assembled causation narrative. Device manufacturers and mobile operators face a subtler calculus — many were counterparties to the condemned agreements and profited from revenue shares — but those squeezed by the conditions rather than enriched by them have theories available.

The obstacles are real and should temper the more breathless projections. Quantifying counterfactual market shares from 2011-2018 in a market that no longer exists is expert-witness warfare of the most expensive kind. Limitation periods, though generally suspended during Commission proceedings and appeals, vary in their national details. Causation — separating harm done by the condemned conduct from harm done by Google simply being better-resourced — is the battleground where these cases are won or lost. And Google settles strategically: quiet resolutions with the strongest claimants, litigation to exhaustion against the rest.

Still, the structural shift is one-directional. Every euro of exposure now sits on the claimant-friendly side of a final judgment, litigation funders openly treat confirmed EU decisions against Big Tech as an asset class, and the Android findings are broader than Shopping’s — covering not one vertical service but the distribution machinery of mobile search itself. The Commission’s fine closed one ledger on 2 July 2026. The private one opened the same morning.

The Digital Markets Act and the rules that now bind Android

The Android judgment adjudicates the past; the Digital Markets Act governs the present, and the interplay between them is where European platform regulation now actually happens. The DMA, applicable in full since March 2024, abandons the case-by-case logic of Article 102 for a list of standing obligations imposed on designated gatekeepers — companies controlling core platform services above size thresholds. Alphabet was designated in the first wave, with Android, Google Search, Chrome, Google Play, YouTube, Google Maps and its advertising services all listed as core platform services.

The DMA reads, in substantial part, like the Android decision converted into statute. Gatekeepers must allow the uninstallation of pre-installed apps and the changing of defaults. They must permit third-party app stores and sideloading on their operating systems. They must not require developers to use the gatekeeper’s ancillary services as a condition of platform access. They must not tie core platform services together as conditions of one another. They must present choice screens for search engines and browsers. Each of these provisions has a direct ancestor in the conduct condemned in the Android case — the legislature took the Commission’s litigation experience and wrote the remedies into permanent, self-executing law, enforceable with fines of up to ten percent of worldwide turnover, rising to twenty percent for repeated infringements, numbers that dwarf even the Android penalty.

The enforcement record against Alphabet under the DMA is already substantial. In 2025 the Commission found Google non-compliant on self-preferencing in search results and on Play Store anti-steering rules that prevented developers from directing users to cheaper offers outside the store. In early 2026, enforcement reached the frontier the Android case never touched: the Commission ordered Google to remove technical barriers preventing rival AI search assistants from operating freely on Android devices, and to share specified search data with competing providers — obligations aimed at ensuring that the default-capture playbook condemned for the browser era is not simply re-run for the assistant era. Separate scrutiny continues over Play Store rules affecting developers and over allegations that Google’s ranking systems demote certain categories of content, including news.

The 2 July 2026 judgment strengthens this entire apparatus in three distinct ways. Legally, several DMA obligations rest on the same analytical foundations — default effects, tying by contract, ecosystem power — that the Court of Justice has now blessed at the highest level; a gatekeeper challenging those obligations as disproportionate must now argue against confirmed case law rather than contested theory. Institutionally, the ruling vindicates the Commission’s judgment across a decade of contested enforcement, political capital that transfers directly to its DMA decisions. Practically, the judgment’s factual findings — status quo bias, the decisiveness of pre-installation, the failure of Google’s quality explanation — are citable evidence in DMA compliance disputes about whether a given choice screen or default flow genuinely works.

The deeper design question the pairing raises is whether the DMA solves the problem the Android case exposed: time. The Android proceeding took a decade from investigation to final judgment, and the market it sought to protect had changed twice over before the gavel fell. The DMA’s answer is to skip the abuse-proving stage entirely — obligations apply by designation, not after litigation. Its first years suggest the gain is real but partial: compliance disputes, specification proceedings and inevitable court challenges reintroduce delay through the back door. What the Android judgment contributes to that ongoing experiment is a floor. Whatever the DMA’s teething problems, no gatekeeper can any longer argue that the conduct it targets was lawful all along. The highest court in Europe has said otherwise, in the largest case ever brought.

Parallel pressure in the United States and the DOJ search case

The Luxembourg judgment did not fall into a vacuum. It landed in a world where Google’s distribution practices have been condemned, in overlapping terms, by courts on the other side of the Atlantic — a convergence that would have seemed implausible when the EU stood alone against Mountain View in 2018.

The centerpiece is United States v. Google, the Department of Justice’s search monopolization case. In August 2024, Judge Amit Mehta of the federal district court in Washington ruled that Google had illegally maintained its monopoly in general search, resting his liability finding on the same behavioral bedrock as the EU case: the exclusive and default distribution agreements — above all the payments to Apple, running above $20 billion a year, and the Android agreements with manufacturers and carriers — that locked Google into the access points where queries begin. The court’s treatment of default power tracked the European analysis almost paragraph for paragraph: defaults decide, users rarely switch, and Google’s own documents proved the company knew it. The remedies decision that followed in September 2025 stopped short of the structural break the government sought — Chrome was not divested — but barred exclusive default arrangements and ordered search data sharing with qualified competitors, remedies whose family resemblance to the EU’s 2026 DMA orders is unmistakable. Appeals continue, but the liability finding stands as the American mirror of what the CJEU confirmed.

A second American front reached its own finality first. In Epic Games v. Google, a federal jury found in December 2023 that Google’s Play Store and billing arrangements constituted unlawful monopolization, and Judge James Donato’s 2024 injunction ordered Google to open Android to rival app stores and alternative billing — an order the Ninth Circuit upheld in 2025. The Epic case attacked the app store layer of the same architecture the EU’s AFA findings addressed from the operating-system side; between them, the two legal systems have now condemned Google’s Android control stack at every level from OS licensing to in-app payments.

The convergence extends beyond the two big jurisdictions. Japan’s Fair Trade Commission issued a cease-and-desist order against Google in 2025 over Android search default requirements — its first such order against a US technology giant — explicitly citing conduct of the type the EU condemned. The UK’s Competition and Markets Authority, wielding its new digital markets regime, designated Google with strategic market status in search in 2025 and is building conduct requirements around defaults and data. Competition authorities in India, South Korea, Turkey and Brazil have all issued Android-related decisions or commitments over the past half-decade, several borrowing the Commission’s market definitions wholesale. A company that once faced one determined regulator now faces a loose global consensus, with the EU judgment as its most authoritative text.

The transatlantic politics remain jagged. Washington has periodically protested EU enforcement and the DMA as discriminatory targeting of American champions, and the 2026 trade environment gives those protests sharper edges. But the protest collides with the record of America’s own courts. When a US federal judge finds that Google’s defaults foreclosed search competition, the claim that Luxembourg invented the theory for protectionist reasons becomes unsustainable. The more accurate description of 2026 is that two legal systems, applying different statutes to the same facts, reached the same conclusion — and that Google’s remaining refuge is not any court’s disagreement but the slowness and caution of remedies everywhere.

Ripple effects for smartphone makers and mobile operators

For the companies that actually build and sell Android devices, the judgment closes a chapter they have lived inside for fifteen years — first as counterparties to the condemned contracts, then as subjects of the remedial reshuffle, now as potential litigants and negotiating beneficiaries.

The manufacturers’ position in the original case was ambivalent by design. Samsung, Xiaomi, Oppo, and their peers signed the MADAs and AFAs; several also took revenue-share payments. They were, in the Commission’s framing, instruments of the abuse rather than its authors — constrained parties whose commercial dependence on the Play Store left them no realistic alternative to Google’s terms. That framing, now final, cuts two ways for them. It absolves them of co-responsibility. It also documents, in a binding judicial record, that Google’s contractual grip on them was strong enough to constitute abuse, a finding of considerable value the next time licensing terms are negotiated.

The post-2018 remedial regime already changed the commercial mechanics in Europe. Google unbundled Search and Chrome from the Play Store licence in the EEA and introduced a paid licensing structure for its app bundle on devices shipping without its search and browser — a fee reportedly reaching up to $40 per device depending on configuration and country tier, though in practice Google’s separate placement payments for Search and Chrome could offset it, and virtually all manufacturers continued shipping the full Google stack. That outcome is itself instructive: given formal freedom, the economics still pointed the same way, because users expect Google services and manufacturers capture placement revenue by providing them. The DMA has since layered on further obligations — default choice screens, uninstallable apps, third-party store access — that apply regardless of contract.

What changes after 2 July 2026 is bargaining posture and legal option value. Manufacturers renegotiating placement deals now do so against a backdrop where Google’s historical terms are adjudicated abuses and its current terms sit under continuous DMA scrutiny; the threat point in every negotiation has moved toward the device makers. Some may explore follow-on claims — particularly any that can document forgone opportunities under the AFAs, such as abandoned fork projects or blocked Amazon partnerships. Most will monetize the new bargaining strength quietly instead, in the form of better revenue-share percentages and looser placement conditions, which is how such judgments typically cash out in industry practice.

Mobile network operators, the case’s forgotten counterparties, share the same structural position. Several took portfolio revenue-share payments in the 2011-2014 window; all distribute Android devices at scale. Their procurement teams have absorbed the same lesson as the manufacturers’: the era in which Google could dictate the software configuration of European Android devices through take-it-or-leave-it bundles is legally closed, and every configuration choice — default assistant, pre-loaded browser, search widget — is now a negotiable, and billable, line item. For an operator sector starved of growth, the ability to auction home-screen real estate that Google once commanded for free is a small but genuine new revenue stream, delivered by a decade of litigation most of them watched from the sidelines.

Consequences for app developers, browser makers and search rivals

Downstream of the manufacturers sits the population the case was nominally fought for: the companies whose products compete with Google’s on Android devices. Their situation after the judgment divides into three distinct stories.

For rival search providers, the ruling is vindication arriving after the battle. DuckDuckGo, Ecosia, Qwant, Brave Search and the other independents spent the 2010s making exactly the argument the courts have now accepted — that no quality of product overcomes a rival’s factory default — and spent the 2020s discovering that remedies designed by the infringer underdeliver. Google’s first choice-screen implementation in 2020 auctioned slots to the highest bidders, a design so hostile that the Commission pressured Google into a free, eligibility-based version in 2021; even then, measured share shifts stayed in the low single digits. The DMA’s stricter choice-screen requirements and the 2026 orders on search data sharing and AI assistant access are the current battleground, and the judgment’s confirmed findings on status quo bias are the rivals’ best ammunition in every compliance dispute — judicial fact, not advocacy, that a default is a market outcome. The realistic prize has also shifted: with search itself being reshaped by AI assistants, the independents’ fight is less about winning the old query box than about preventing its successor from being captured the same way.

For browser makers, the Chrome tying finding lands in a market Chrome already won. Opera, Mozilla’s Firefox, Vivaldi and newer entrants like Brave compete on Android against a browser that shipped on effectively every device for a decade; Chrome’s mobile share in Europe remains commanding, and no judgment un-ships those defaults retroactively. The value of the ruling to this group is prospective and litigious: prospective, because DMA browser choice screens on Android — rolled out from 2024 — must now be judged for effectiveness against the case’s findings rather than Google’s assurances; litigious, because a final tying finding covering Chrome’s distribution from 2012 to 2018 is the predicate for damages theories that Opera and Mozilla, both of which also depend on Google search payments in an awkward twist, have so far pursued cautiously if at all. The dependency point is worth dwelling on — Mozilla’s revenue is dominated by Google’s default payments in Firefox — because it illustrates how thoroughly the condemned model wired itself into even its victims’ business plans.

For app developers, the judgment’s effects are real but indirect, flowing through the app store layer rather than the search layer. The AFA findings and their statutory descendants in the DMA underwrite the opening of Android to alternative distribution: third-party app stores, direct sideloading with fewer friction screens, alternative billing without Google’s 15-to-30-percent commission. Combined with the Epic v. Google injunction in the United States, the direction across jurisdictions is uniform, and the commercial stakes are enormous — Play Store fees are one of Alphabet’s most profitable revenue lines, and every percentage point of billing that escapes them transfers directly to developer margins. European developers of subscription apps, games and marketplaces are the concrete beneficiaries: they can steer users to their own payment flows, list in rival stores, and cite a final judgment whenever Google’s compliance turns grudging. The judgment does not name them, but they are the residual claimants of everything it decided.

Across all three groups runs a shared, sobering lesson about time. The businesses best positioned to exploit the judgment are largely not the ones harmed by the conduct; many of those shrank, pivoted or died between 2011 and 2026. Competition enforcement in platform markets, even when it wins completely, compensates the market’s future occupants rather than restoring its past ones — which is precisely the argument European legislators used to justify moving from litigation to standing regulation.

Stakes for the advertising market and SEO practitioners

The Android case is, at bottom, an advertising case — the contracts existed to protect the query stream that feeds the auction that funds Alphabet — and its final resolution carries practical weight for everyone whose business depends on visibility in Google’s results, from media buyers to search marketers.

Start with the advertising market structure. Google’s share of European search advertising remains dominant in 2026, and nothing in the judgment changes that directly; fines do not reallocate budgets. What the judgment changes is the regulatory trajectory around the edges of that dominance. The confirmed doctrine strengthens the Commission’s hand in the still-open adtech proceedings — the separate investigation into Google’s advertising technology stack, where the Commission has found self-preferencing across the buy side, sell side and exchange, imposed a €2.95 billion fine in 2025, and continues to weigh structural remedies including divestiture. It likewise reinforces DMA obligations on advertising transparency and data combination. Media buyers should expect the gradual, enforced opening of measurement, data access and placement mechanics that Google historically kept closed, with each opening creating small arbitrage opportunities for those paying attention before the rest of the market reprices.

For SEO and content businesses, the more consequential thread is the one running from the judgment’s default findings to the 2026 orders on AI search. The economic logic that made the search box worth €4 billion of legal risk applies with equal force to AI assistants and answer engines: whoever owns the default interface owns the query, and the query is the monetizable moment. European regulators have said explicitly that they intend to prevent a re-run — hence the orders requiring Android to accommodate rival AI assistants and requiring search data sharing that could feed competing answer engines. If those remedies bite, the practical consequence for visibility professionals is a slow diversification of the surfaces that matter: optimization for a single ranking system gives way to optimization across several retrieval and answer systems, each with its own citation behavior. Businesses that treat generative engine optimization as a sibling discipline to classic SEO are positioning for the market structure European law is trying to force into existence.

There is also a defensive reading for publishers and site owners. The judgment’s confirmation that Google’s distribution advantages were built by contract rather than earned purely on merit feeds directly into adjacent European disputes — over the treatment of news content in results, over AI Overviews’ effect on referral traffic, over data used to train and ground Google’s generative features. Publisher complaints on several of these fronts are pending before the Commission, and each arrives now before an institution whose theory of Google has just been ratified at the highest level. None of this guarantees outcomes; it shifts priors. For a decade, the safe assumption in the visibility industry was that Google’s architecture was the weather — permanent, unappealable, to be adapted to. The Android judgment’s deepest commercial message is that the weather is now partly man-made in Brussels, and forecasting it requires reading regulators as closely as reading algorithms.

Android in 2026 with choice screens, forks and life after remedies

Judged against the market it condemned, the judgment describes a lost world. The Android of 2026 differs from the Android of 2018 in nearly every respect the case touched, and taking stock of those differences is the honest way to measure what enforcement achieved and what it did not.

The contractual layer changed first and most completely. The MADA bundle condemned in the decision no longer operates in its condemned form in Europe: Search and Chrome are unbundled from the Play Store licence, the exclusivity payments are gone, the anti-fragmentation commitments were reworked into arrangements that formally permit incompatible devices. The DMA then hard-coded the remedial state: choice screens for search and browser at device setup, uninstallable Google apps, third-party app store access, sideloading with reduced friction. On paper, a European Android buyer in 2026 faces more configurable choice at first boot than any mainstream computing platform has ever offered.

The market layer changed far less. Google Search’s European share remains above ninety percent on mobile. Chrome remains the dominant mobile browser. The Play Store remains the app economy’s center of gravity, its rivals — Samsung’s Galaxy Store, Epic’s store, F-Droid, regional players — collectively marginal. Choice screens moved shares by points, not tens of points. The gap between restored formal competition and unmoved market outcomes is the case’s defining empirical result, and it admits two readings that will contest each other in policy debates for years. Google’s reading: users, offered genuine choice, chose Google, proving quality drove the shares all along. The enforcers’ reading: fifteen years of default capture built habits, data advantages and ecosystem lock-in that formal choice arriving in the 2020s could not dislodge — the harm compounded beyond the reach of remedy, which is an argument for earlier intervention, not weaker findings.

The fork question resolved in its own ironic way. The AFA findings freed manufacturers to build incompatible Android variants, and European market forces produced essentially none; the one fork that achieved scale, Huawei’s HarmonyOS, was created by American export sanctions, matured in China, and treats Europe as a secondary market. Meanwhile the operating-system duopoly the case implicitly worried about — Android and iOS, with no third force — is more entrenched in 2026 than in 2018. Whatever window existed for a fork-based challenger, the litigation record suggests it closed around the middle of the last decade, while the agreements that helped close it were still being investigated.

The frontier moved, too. The strategic layer of a smartphone in 2026 is decreasingly the search widget and increasingly the assistant — the AI layer that answers directly, books directly, and mediates the apps beneath it. Google’s Gemini occupies that layer on Android by default; the Commission’s 2026 DMA orders on rival assistant access exist precisely because regulators recognize the pattern repeating one level up the stack. The Android judgment thus closes as its subject matter reopens in new form: the specific contracts are history, the strategic logic they served — own the default, own the user — is the live question of the assistant era, now confronted with better law, faster tools and a fully briefed regulator.

Android case milestones from complaint to final judgment

The table compresses thirteen years into ten rows, and its most eloquent feature is the spacing: five years from complaint to decision, eight more from decision to finality. Every structural argument about platform regulation in Europe — for the DMA, for interim measures, for ex-ante rules — is ultimately an argument about those intervals.

Lessons for platform businesses operating in Europe

Beyond Google, the judgment is a compliance document for every company that runs a platform, and its operational lessons can be stated concretely enough for a board memo.

The first lesson concerns bundles. A dominant platform component — an app store, an operating system, a marketplace, an API without which partners cannot function — cannot be used as consideration for the distribution of adjacent products. The Android tying analysis makes the trigger conditions explicit: distinct products, a must-have tying product, conditions partners cannot realistically refuse, and a distribution advantage rivals cannot replicate through ordinary competition. Crucially, zero pricing is no defense; the courts will treat placement, defaults and data as the currency in which free products are paid for. Any dominant firm whose partner contracts contain the words “must pre-install,” “must set as default” or “as a condition of licensing” should read them this week against the judgment.

The second lesson concerns ecosystem rules. Platform stewards may impose compatibility, security and quality requirements — the judgment reaffirms the legitimacy of protecting an ecosystem where the steward’s products operate. The boundary is functional: restrictions must protect something real, and restrictions whose only demonstrable effect is preventing rivals or forks from existing will be severed and condemned, with justifications assessed clause by clause. Drafting discipline follows: every restrictive term in a platform agreement should trace to a documented, proportionate integrity interest, written down at the time, because litigation will excavate the contemporaneous record — as it excavated Google’s internal documents on the value of defaults.

The third lesson concerns strategy documents themselves. Across the Android, Shopping and US proceedings alike, the most damaging evidence was Google’s own: internal analyses proving the company understood default power precisely as the regulators alleged. Platforms should assume that every deck quantifying the value of a default, every email describing a contract as protecting share, will be Exhibit A a decade later. That is not a counsel of concealment — destruction and evasion carry their own catastrophic risks, as Google’s chat-retention practices in US litigation showed — but of alignment: conduct that cannot be described honestly in internal documents without sounding like foreclosure probably is foreclosure.

The fourth lesson concerns time horizons and venue. The Android matter ran thirteen years from complaint to finality, and the exposure it created runs forward from here through damages litigation and DMA enforcement. European regulatory risk is not an event; it is a decades-long balance-sheet item, senior to most strategic bets, and it compounds — each confirmed finding feeding the next proceeding, each fine raising the baseline for recidivism multipliers. Companies designing global products around a single integrated architecture should price the possibility that Europe will require a structurally different version, because it now regularly does. The firms that handle this best treat Brussels not as a litigation adversary but as a second product-requirements authority, and they staff accordingly.

Open questions the judgment leaves unresolved

Final judgments settle cases, not debates, and intellectual honesty requires cataloguing what the 2 July 2026 ruling does not answer.

The largest open question is remedial effectiveness. The judgment confirms that Google’s conduct was illegal; it cannot establish what the European search and mobile markets would look like had the conduct never occurred, or whether any remedy imposed in the 2020s can meaningfully reconstruct the competition foreclosed in the 2010s. The stubborn stability of Google’s market shares after unbundling and choice screens is consistent both with the merit story Google tells and with the compounded-harm story enforcers tell, and no methodology can now cleanly separate them. That ambiguity will be exploited by both sides in every future proceeding, and it is genuinely unresolved.

A second question is doctrinal reach. The judgment’s loosening of the as-efficient-competitor requirement and its blessing of context-wide effects analysis were crafted against a defendant of unique scale. Lower courts and national authorities must now decide how far those holdings travel — whether they apply to every dominant firm or only to super-dominant, multi-market platforms whose position no hypothetical rival could replicate. Read broadly, the judgment lowers the intervention threshold across the whole economy; read narrowly, it creates a special regime for a handful of gatekeepers. The text supports either reading, and the boundary will be litigated for a decade.

A third question concerns the assistant era directly. The judgment adjudicates search boxes and browsers; the strategic surface has moved to AI assistants, where the relevant inputs — models, compute, data, device integration — are distributed differently than they were in 2011. Whether the default-capture framework transfers cleanly to a world where the assistant is deeply integrated into the operating system for genuine functional reasons, and where the marginal competitor is another trillion-dollar firm rather than a startup search engine, is untested. The Commission’s 2026 orders on assistant access are the first experiment; their court challenges will be the sequel to this case.

Smaller but consequential uncertainties round out the list. The scale of follow-on damages exposure is unknowable in advance — PriceRunner suggests the ceiling is high, but each claim turns on causation battles the infringement finding does not resolve. The interaction between Article 102 case law and DMA enforcement remains institutionally unsettled: whether the DMA displaces, supplements or merely accelerates classical enforcement will be defined by the first major conflicts between the two tracks. And the geopolitical variable is live: technology regulation now moves inside a broader transatlantic negotiation over trade and security, and the durability of Europe’s enforcement posture under sustained political pressure from Washington is a question no court can answer. The judgment gives the Commission maximum legal authority at a moment when the constraints on using it are increasingly political rather than legal.

The road ahead for Google in Brussels

For Google, 2 July 2026 closes the longest front of its European war and clarifies the map of the remaining ones. The company’s Brussels agenda for the coming years is legible, and none of it is comfortable.

The immediate docket is dense. The adtech case remains the largest live classical proceeding: the Commission’s 2025 decision found self-preferencing across Google’s advertising stack, imposed a €2.95 billion fine, and kept structural separation — forced divestiture of parts of the stack — explicitly on the table, with the follow-up phase running through 2026. DMA enforcement proceeds in parallel on multiple tracks: compliance with the non-compliance findings on search self-preferencing and Play Store steering, implementation of the 2026 orders on AI assistant access and search data sharing, and continuing scrutiny of ranking practices affecting publishers. Each track carries penalty exposure calibrated to worldwide turnover, and each now unfolds against a judicial backdrop in which Google’s core defenses — quality explains our position, our contracts merely reflect merit, effects must be proven counterfactually — have been examined and rejected at the highest level.

The private-litigation front will grow for years. The Android findings’ breadth invites claims from search rivals, browser makers and app store operators across multiple member states, with litigation funders already active; PriceRunner established the order of magnitude national courts will countenance. Google’s rational strategy — settle the strongest, exhaust the weakest — is expensive in both directions, and every settlement quietly reprices the next claim.

Strategically, the deeper question is whether Google changes its European operating model or continues to litigate its way through a permanent state of enforcement. The record to date suggests the latter: compliance implemented at the minimum defensible level, remedies engineered to preserve as much of the status quo as survives review, and appeals taken to the last instance. That approach maximized short-term revenue for a decade and has now produced its full bill — roughly €11 billion in fines, binding adverse findings across three cases, statutory obligations written from its own conduct, and a legal environment in which Google enters every new proceeding as an adjudicated recidivist. The alternative model, visible in fragments when Google over-complies to preempt designation fights, would treat Europe as a jurisdiction whose rules are inputs to product design rather than costs to be minimized. Which model prevails inside the company is a governance question the judgment sharpens but cannot decide.

What can be said with confidence is that the era this judgment closes will not return. The Android playbook — capture the default, contract the ecosystem shut, monetize the funnel, litigate the consequences for a decade — was the most successful distribution strategy in the history of software, and it is now, in Europe, both illegal in its past form and pre-empted in its future ones. The Court of Justice needed one sentence to end eight years of argument: the appeal is dismissed. Everything Google builds in Europe from here will be built inside that sentence’s shadow, and so, more consequentially, will everything built by the platforms that watched Google test the boundaries and lose.

Android’s origins and the mobile operating system wars that shaped the case

The conduct condemned in the judgment makes full sense only against the industrial history that produced it, and that history begins before the iPhone. Google acquired Android Inc., a fifty-person startup led by Andy Rubin, in 2005 for a reported figure around $50 million — in retrospect among the most consequential acquisitions in technology history. The strategic motivation was defensive from the first day: Google’s leadership feared a mobile future controlled by Microsoft, whose Windows Mobile was then the presumed heir to the desktop monopoly, and in which Google’s search would sit behind a rival’s gate. Android was insurance against a world where someone else owned the door to the internet.

The launch of the iPhone in 2007 turned the insurance policy into an offensive weapon. Apple demonstrated what the smartphone would be; Google’s response, formalized through the Open Handset Alliance in late 2007 and the first Android device in 2008, was to give the rest of the industry a way to build competitive hardware without writing an operating system. The incumbents Android displaced tell the story of why manufacturers accepted Google’s terms so readily. Nokia’s Symbian, the world’s dominant smartphone platform through the late 2000s, collapsed under its own fragmentation and developer hostility — the cautionary tale Google would later invoke to justify the anti-fragmentation agreements. BlackBerry clung to hardware keyboards and enterprise email while the market moved to touchscreens and apps. Microsoft, the feared hegemon, fumbled the transition entirely: Windows Mobile was scrapped, Windows Phone launched late in 2010, and despite acquiring Nokia’s device business for over €5 billion in 2014, Microsoft never escaped single-digit share and abandoned the platform. By roughly 2013, the plausible mobile futures had collapsed to two: Apple’s closed garden and Google’s conditionally open one.

That collapse is the hidden premise of the Commission’s market definitions. When the decision defines a market for licensable smart mobile operating systems and finds Android above ninety-five percent of it, the finding encodes the graveyard: Symbian dead, Windows Phone dying, BlackBerry irrelevant, Samsung’s Tizen and the Firefox OS experiments stillborn. A manufacturer in 2014 wanting to sell a competitive smartphone had exactly one operating system available for licence, and the owner of that system knew it. The MADAs, AFAs and RSAs were drafted into that position of strength. Google’s litigation argument that Apple constrained it was, in the manufacturers’ lived reality, beside the point — Samsung could not respond to harsh Google terms by licensing iOS.

The history also explains the timing of the infringement period. The Commission dated the abuse from January 2011 — the moment, roughly, when Android’s victory in the licensable market became irreversible and its terms hardened accordingly. Early Android had been genuinely permissive; the contractual tightening tracked the disappearance of alternatives. That sequence, from openness while contestable to restriction once entrenched, is the pattern competition lawyers now teach from this case, and it supplies the answer to the perennial defense that Android’s openness disproves the abuse: the openness was real, was strategic, and was progressively qualified precisely where it mattered once the war was won.

One more historical thread bears on the judgment’s meaning. The fear that birthed Android — a rival controlling the gateway to Google’s economics — never disappeared; it migrated. It drove the Apple default payments condemned in the American case, the browser strategy behind Chrome, and today the integration of Gemini into every Android surface. The judgment condemns one decade’s expression of a permanent strategic anxiety. The anxiety survives its condemned instruments, which is the deepest reason regulators treat the assistant era with such suspicion: the company’s incentive structure has not changed, only its tools.

Echoes of Microsoft and the lineage of European tying law

Veterans of European competition law experienced the Android case as a sequel, because in its basic architecture it replays the Commission’s confrontation with Microsoft two decades earlier — with the roles reversed in an irony no one in the proceedings failed to notice.

The Microsoft case of the 2000s established the modern European law of platform tying. The Commission’s 2004 decision condemned Microsoft for bundling Windows Media Player with Windows and for withholding interoperability information from workgroup-server rivals, imposing a then-record €497 million fine; the Court of First Instance upheld it comprehensively in 2007, and a follow-on decision in 2009 addressed Internet Explorer’s bundling with Windows, resolved through commitments that produced the first browser choice screen — the direct ancestor of the screens Android users see today. The doctrinal machinery built for Microsoft — distinct products, must-have tying product, distribution advantage no rival can match, foreclosure through ubiquity rather than price — is the machinery the Commission took off the shelf for Android. Even the evidence rhymed: in both cases, the decisive material included the defendant’s internal documents describing distribution as the weapon.

The ironies are instructive rather than decorative. Google was among Microsoft’s loudest accusers in the browser-tying era, arguing that default distribution of Internet Explorer foreclosed competition on the merits — the precise theory later turned on Chrome. Microsoft, for its part, appeared in the Android story on the other side of the table, as a FairSearch participant and as the operator of the Windows Mobile platform whose Bing-default devices supplied the Commission’s cleanest natural experiment. The two cases together establish that European tying law is symmetric: it condemned the desktop incumbent to the benefit of, among others, Google, and then condemned Google’s mobile incumbency by the same measure. That symmetry is the strongest available answer to the charge that EU enforcement targets American firms as such; it targets dominant distribution, and has done so consistently across twenty-five years and changing defendants.

The comparison also measures what changed between the eras. Microsoft’s fine was half a billion euros; Google’s is eight times larger, reflecting both the fining guidelines’ evolution and the vastly greater revenues at stake. Microsoft’s case took roughly six years from decision to final judgment; Google’s took eight, despite every institutional effort at acceleration — evidence that judicial timelines resist reform even as markets accelerate. And the remedial lesson darkened: the Windows Media Player unbundling famously failed (users ignored the stripped version), the browser choice screen modestly succeeded, and Android’s remedies have so far tracked the Media Player precedent more than the browser one. European legislators absorbed all three data points; the DMA’s designers cite the Microsoft-to-Android arc explicitly as proof that case-by-case tying enforcement, however doctrinally sound, arrives too late to restore what bundling forecloses — the argument that moved Europe from litigating platform conduct to legislating it.

For Google, the Microsoft parallel carries one final, uncomfortable projection. Microsoft’s European reckoning coincided with the end of its most imperial phase; constrained at home and abroad, distracted by compliance, it missed mobile — and its eventual renaissance came from businesses, cloud and enterprise software, far from the condemned bundle. Whether the Android judgment marks a similar inflection for Google — the moment enforcement, private litigation and a platform shift toward AI jointly loosen a hegemony that law alone never could — is the historical question the next decade will answer.

The Shopping and AdSense cases and the full map of Google’s EU liability

The Android judgment is one panel of a triptych, and its meaning sharpens when read alongside the fates of the two sibling decisions, because together they define exactly which parts of Google’s conduct European law has condemned, which it has excused, and which remain in play.

The Google Shopping case came first and traveled furthest. The Commission’s June 2017 decision found that Google had abused its search dominance by systematically favoring its own comparison-shopping service in general results while demoting rivals through ranking algorithms, and imposed a fine of €2.42 billion. The General Court upheld the decision in substance in November 2021, and the Court of Justice dismissed Google’s final appeal in September 2024 — making Shopping the first of the trilogy to reach finality and establishing self-preferencing by a dominant platform as a standalone abuse under Article 102, a doctrinal innovation with consequences far beyond comparison shopping. The Shopping finding is the predicate for the PriceRunner damages award and for a queue of comparison-service claims across Europe; it is also the conceptual ancestor of the DMA’s self-preferencing ban and of the live adtech and search-ranking proceedings. Where Android condemned how Google acquired queries, Shopping condemned what Google did with the results page once it had them.

The AdSense case closed the circuit on the money side — and broke the Commission’s winning streak. The March 2019 decision, the last of the Vestager trilogy, fined Google €1.49 billion for exclusivity and premium-placement clauses in its agreements with large website publishers using AdSense for Search, clauses the Commission said kept rival search-advertising intermediaries off publishers’ sites between 2006 and 2016. In September 2024, the General Court annulled the decision in its entirety: it agreed that the clauses could restrict competition in principle but found the Commission had failed to properly establish the duration and coverage needed to demonstrate the foreclosure effect, and had not adequately considered all relevant circumstances. The Commission accepted the defeat rather than appeal. AdSense thus stands as the trilogy’s proof of judicial independence — and as a recurring exhibit for Google’s lawyers, who cite it whenever the Commission’s economic analysis is challenged, exactly as they cite the Android RSA annulment.

Read as a set, the three outcomes draw a precise doctrinal map. Structural and architectural theories prevailed: tying (Android), anti-fragmentation (Android) and self-preferencing (Shopping) are confirmed abuses, established largely through the design and documented purpose of Google’s systems and contracts. Payment-and-clause theories requiring quantified foreclosure struggled: the portfolio RSAs (Android) and the AdSense exclusivity clauses both fell to evidentiary standards the Commission failed to meet. The Commission has visibly internalized the pattern — its subsequent adtech decision leads with self-preferencing architecture rather than contractual economics, and its DMA obligations are drafted as structural rules that require no effects proof at all.

The financial ledger of the trilogy also frames the Android judgment’s proportions. Of roughly €8.25 billion originally imposed across the three decisions, approximately €6.55 billion survived judicial review — €2.42 billion from Shopping and €4.125 billion from Android — while €1.49 billion from AdSense and €218 million from the Android reduction were returned or never collected. A survival rate near eighty percent, across the most aggressively litigated defense effort in the history of competition law, is the number that tells national authorities, claimants and gatekeepers alike how much of what the Commission writes ultimately becomes law.

Practical steps for European businesses acting on the ruling

A judgment of this scale generates obligations and opportunities that differ sharply by position in the ecosystem, and readers who act on news rather than merely follow it can translate the ruling into a short, concrete agenda.

For businesses that believe they were harmed — search services, browser developers, app distributors, comparison and vertical-search operators active between 2011 and 2018 — the first step is a limitation-period audit. National limitation rules for follow-on damages typically run from the moment the infringement decision becomes final, which the CJEU ruling has now fixed at 2 July 2026, but the details vary by member state and by when the claimant knew of the harm; a claim allowed to lapse is worth nothing. The second step is evidence preservation and reconstruction: traffic data, distribution negotiations that failed, internal analyses of Android placement economics from the infringement period. The third is a funding conversation — litigation funders actively underwrite follow-on claims against confirmed EU decisions, converting a legal position into a financeable asset, and the PriceRunner award has repriced the category upward. None of this commits a company to suing Google, with whom most potential claimants also do business; it preserves the option, which after this judgment has measurable value.

For companies negotiating with Google — device makers, operators, large publishers, major app developers — the ruling is a bargaining chip to be spent deliberately. Contract renewals touching placement, defaults, revenue shares or store terms should be benchmarked against both the condemned practices and current DMA obligations, with counsel mapping every clause that resembles the adjudicated conduct. Complaints to the Commission under the DMA are cheaper and faster than Article 102 complaints ever were, and the credible threat of one now carries the weight of a fully vindicated enforcement record.

For platform operators of any size approaching dominance in their own niche — marketplaces, SaaS ecosystems, vertical app stores — the agenda is defensive. A focused audit should test three exposures the judgment defines: whether any must-have product is contractually bundled with adjacent products; whether ecosystem rules restrict partners’ conduct beyond the footprint of the platform’s own services; and whether internal documents describe defaults or placement in the vocabulary of foreclosure. Remediation is radically cheaper before a complaint than after one, and the judgment’s clause-by-clause proportionality method gives compliance teams a usable template: every restriction must trace to a documented integrity interest, or it is a liability in waiting.

For advertisers, SEO professionals and publishers, the actionable horizon is the remedial one. Choice screens, assistant-access orders and search-data-sharing obligations will incrementally diversify the surfaces where European users begin their sessions; visibility strategies that track regulatory implementation — which rivals gain slots, which data flows open, how AI answer engines source citations — will capture the early arbitrage that always accompanies enforced market opening. The professionals who treated the 2018 choice screens as a monitoring project rather than a footnote gained little, because the remedy underdelivered; the 2026 generation of remedies is stronger, and the same monitoring discipline costs little to maintain against the possibility that this time the shares move.

The common thread across all four postures is that the judgment converts uncertainty into structure. For eight years, every plan touching Google’s Android conduct carried a contingency: perhaps the courts disagree. That contingency is retired. Planning against a settled legal fact is a different and easier exercise, and the businesses that update fastest extract the most from the interval before the new equilibrium becomes common knowledge.

Privacy, data power and the dimension the fine never measured

The Android case was prosecuted as a competition matter, but its subject — the machinery that routed billions of users’ queries to a single company — is equally a data story, and the judgment’s silence on that dimension is itself worth analysis.

The condemned contracts did more than protect advertising revenue; they guaranteed Google an unmatched behavioral data harvest during the years when mobile data became the raw material of machine learning. Every default search box on every European Android device fed queries, locations, click patterns and session behavior into systems that improved ranking, ad targeting and, eventually, the foundation models behind Google’s AI products. Competition economists recognize this as the data feedback loop: scale produces data, data produces quality, quality justifies the scale — a cycle that converts a distribution advantage into a permanent product advantage even after the distribution contracts are dismantled. The stubborn post-remedy stability of Google’s market share, the case’s central empirical puzzle, is at least partly this loop operating as designed; rivals given equal shelf space in 2024 still compete against fifteen years of accumulated training signal they can never retroactively collect.

European law has begun to treat that loop as a regulatory object in its own right. The DMA prohibits gatekeepers from combining personal data across core platform services without consent, mandates data portability, and — in the search context specifically — requires gatekeepers to share ranking, query, click and view data with rival search engines on fair terms, the obligation the Commission’s 2026 orders against Google operationalize. The German competition authority’s landmark Facebook case had earlier established that data-exploitation terms can themselves be an abuse of dominance, and the Court of Justice confirmed in 2023 that competition authorities may consider GDPR compliance within abuse assessments. The intellectual traffic between the two fields runs in both directions: privacy regulators cite market power to explain why consent is illusory, competition regulators cite data accumulation to explain why dominance persists.

The Android judgment contributes to this convergence indirectly but materially. Its confirmed findings on status quo bias describe the exact mechanism through which data concentration happened — users did not choose to send fifteen years of queries to one company; the architecture chose for them. Privacy advocates read that finding as judicial confirmation that the consent framework of the same period was built on defaults users never meaningfully accepted. Data-protection authorities and claimants in GDPR litigation now possess, courtesy of a competition judgment, a top-court factual record about how default architecture governs user behavior — a record likely to surface in proceedings about consent design, dark patterns and the legitimacy of pre-ticked ecosystems for years.

The unmeasured dimension also frames the forward-looking risk honestly. The fine priced the advertising revenue the conduct protected; no instrument priced the data advantage it accumulated, and that advantage is precisely what powers Google’s position in the AI transition the regulators now police. If the assistant era entrenches along the lines the search era did, the retrospective judgment on the Android case may be that Europe fined the symptom, remedied the contracts, and never touched the asset — which is the strongest argument circulating in Brussels for remedies that address data directly, up to and including the structural options still on the table in the adtech proceeding.

Frequently asked questions about the Google Android ruling and the €4.1 billion fine

Author:
Jan Bielik
CEO & Founder of Webiano Digital & Marketing Agency

This article is an original analysis supported by the sources cited below

Google Android: the Court of Justice upholds Google’s fine of around €4.1 billion Official press release No 93/26 of the Court of Justice of the European Union, 2 July 2026, announcing the dismissal of the appeal in case C-738/22 P and the confirmation of the penalty.

Judgment of the Court (Second Chamber) of 2 July 2026, Google and Alphabet v Commission Full text of the CJEU judgment in C-738/22 P, including the fining methodology, the treatment of each ground of appeal and the operative part.

C-738/22 P Google and Alphabet v Commission, case documents InfoCuria case file for the appeal, collecting the appeal notice of 30 November 2022, procedural orders, the Advocate General’s opinion and the judgment.

Antitrust: Commission fines Google €4.34 billion for illegal practices regarding Android mobile devices The European Commission’s press release of 18 July 2018 setting out the original decision, the three categories of restrictions and the fine calculation.

Statement by Commissioner Vestager on the Google Android decision Margrethe Vestager’s statement of 18 July 2018 explaining the Commission’s reasoning on pre-installation, anti-fragmentation and revenue share payments.

Case AT.40099 Google Android, European Commission competition case file The Commission’s official case page containing the decision documents, summaries and procedural history of the Android proceeding.

The General Court largely confirms the Commission’s decision that Google imposed unlawful restrictions Press release No 147/22 on the General Court’s judgment of 14 September 2022 in T-604/18, covering the partial annulment and the fine reset at €4.125 billion.

Advocate General Kokott proposes that the Court of Justice dismiss Google’s appeal Press release No 72/25 of 19 June 2025 summarizing the Advocate General’s opinion recommending full dismissal and confirmation of the fine.

Opinion of Advocate General Kokott of 19 June 2025 in Case C-738/22 P Full text of the opinion, including the analysis of counterfactual requirements and the as-efficient-competitor benchmark in digital markets.

Google loses fight against record €4.1 billion EU antitrust fine Reuters report from Brussels on the judgment, its context within fifteen years of EU enforcement against Google and the reaction of the parties.

Google Loses EU Court Appeal Over €4.1 Billion Android Antitrust Fine Bloomberg’s coverage of the ruling, including complainant FairSearch’s reaction and the Commission’s response to the court win.

Google loses legal fight over 4.1 billion-euro EU antitrust fine CNBC report covering the market reaction, Alphabet’s premarket share move and Google’s statement on Android openness and choice.

Google loses final appeal to overturn $4.1 billion EU fine Technology press account of the judgment focusing on the contractual restrictions condemned and the end of the appeal road.

Google’s appeal fails as EU’s top court upholds €4.1 billion Android fine Android-focused coverage explaining what the ruling means for the Android licensing model and for users of the platform.

Google loses final appeal over Android’s €4.1 billion EU antitrust fine Report placing the judgment in the context of Google’s cumulative EU penalties and the parallel obligations arriving under the Digital Markets Act.

EU Court of Justice upholds €4.1 billion Google Android fine EU affairs monitoring service reproducing the court’s findings in detail, including the revised joint and several liability of Alphabet.

The EU Court of Justice upholds a Big Tech company’s €4.125B fine in the mobile ecosystem Competition law bulletin entry from Concurrences collecting scholarly commentary on the judgment and the preceding opinion.

Google Android Case C-738/22 P: A Step Toward Final Judicial Resolution Law firm analysis of the six grounds of appeal and the significance of the Advocate General’s reasoning on the as-efficient-competitor test.

General Court Partially Annuls European Commission Decision in Google Android Cleary Gottlieb’s detailed client analysis of the 2022 General Court judgment, the market definitions and the annulment of the revenue share findings.

The Recognition that Dare Not Speak Its Name: AG Kokott’s Opinion in Google Android Academic op-ed on EU Law Live examining the doctrinal shifts in the opinion on counterfactual analysis and the equally efficient competitor benchmark.

Google told to pay record $1.5bn damages to Klarna’s PriceRunner by Swedish court MLex report on the Stockholm Patent and Market Court’s award of 1 July 2026, the largest antitrust damages ruling in Swedish history.

Swedish court says Google is to pay $1.5 billion to Klarna in antitrust damages Reuters account of the PriceRunner follow-on damages ruling flowing from the Google Shopping decision, delivered the day before the Android judgment.

Google LLC v Commission, case overview Encyclopedic summary of the Android litigation with references to the 2013 FairSearch complaint, the 2015 opening of proceedings and the case documents.

Google loses €4.1bn EU fight over the apps Android users see first European coverage connecting the final judgment to the April 2026 preliminary findings the Commission sent Google under the Digital Markets Act.