Google's €4.1 billion Android fine is final. The fine was never the part that binds.

The reported version of Wednesday's ruling is simple and slightly boring: Google lost, again, and will pay a fine of about €4.1 billion for what it did with Android. That is true, and it is the least consequential sentence you could write about the decision. Google provisioned for this money years ago; €4.1 billion is a rounding error against Alphabet's cash, and the conduct being punished ended in 2018. If the fine were the whole story, the correct reaction would be a shrug. The reason to pay attention is not the number the Court of Justice confirmed on 2 July. It is the door the confirmation opens — a door that has nothing to do with the European Commission's bank account and everything to do with a directive most people have never heard of.

Let me take it in the order that actually matters: what the Court did, what the finding now legally is, and who can pick it up and use it. The fine is the headline. The bindingness is the story.

What the Court actually decided

The case is C-738/22 P, and the "P" matters: it marks an appeal to the Court of Justice, the highest court in the European Union, from which there is no further appeal. This is the end of the road. In 2018 the Commission fined Google €4.34 billion for abusing the dominance of Android — an operating system that runs on the large majority of the world's phones — in three distinct ways. It required manufacturers to pre-install Google Search and the Chrome browser as a condition of getting the Play Store, the app store no competitive Android phone can ship without. It paid some large manufacturers and network operators to install Google Search exclusively. And it blocked manufacturers that wanted to sell any Google-approved device from also selling devices running rival, unapproved versions of Android — the "anti-fragmentation" agreements that kept the alternatives from ever reaching a shelf.

In 2022 the General Court, the EU's second-highest, largely upheld the Commission but trimmed the fine to €4.125 billion, recalculating one element. Google appealed that to the Court of Justice. On 2 July 2026 the Court of Justice dismissed the appeal in its entirety. The operative words are the ones that travel: the Court "dismisses the appeal brought by Google and Alphabet," thereby confirming the penalty. There is no clause sending anything back for reconsideration, no partial remittal, no invitation to re-plead. The finding of infringement is now definitive across the Union.

The fine is the headline. The bindingness is the story. One is a number Google wrote off years ago; the other is a fact no rival will ever have to prove again.

Why "definitive" is a legal event, not an adjective

Here is where the summary and the binding text part company. To most readers, "the ruling is final" sounds like a note about mood — the saga is over, everyone can move on. In EU law it is a switch being thrown. Two instruments turn a final infringement finding into a usable weapon in someone else's hands, and they are worth naming because the coverage almost never does.

The first is Regulation 1/2003, the workhorse of EU competition enforcement, whose Article 16 forbids national courts from taking any decision that runs counter to a Commission decision the EU courts have upheld. A judge in Munich or Milan cannot now entertain the argument that Google did not, in fact, abuse Android's dominance. That question is closed, for every court in the Union, simultaneously. The second is Directive 2014/104, the Antitrust Damages Directive, transposed into the national law of every member state. Its Article 9 provides that a final infringement decision is to be treated as irrefutably established for the purposes of a damages claim before the courts of that country — and as at least prima facie proof everywhere else in the EU. Translated out of the drafting: a company that believes Google's conduct cost it money no longer has to prove that the conduct was unlawful. That part is done. It only has to prove two things — that it suffered a loss, and that this loss was caused by the infringement the Court has just sealed.

If you have never litigated one of these cases, it is hard to convey how much of the mountain that removes. In a stand-alone competition claim, establishing the abuse — the market definition, the dominance, the theory of harm, the years of internal documents — is most of the cost, most of the time, and most of the risk. A follow-on claim inherits all of that as settled fact and starts at the part plaintiffs can actually manage: what did it cost me, and can I trace it to you. The eight years Google spent appealing did not just fail to overturn the fine. They built, brick by brick, an unappealable foundation that its competitors can now stand on for free.

Who is now in scope to sue

So the useful question is not "how big was the fine" but "who was harmed by the three specific practices the Court just confirmed, and can they show it." The answer is a longer list than Google would like.

• Rival search engines that were kept off the default position on hundreds of millions of Android phones between 2011 and 2018 — the precise years the Commission's decision covers. Yandex, one of the original complainants in the case, is the obvious name, but any general search competitor that can model the traffic and revenue it lost to the default arrangement has a route.
• Browser makers that could not displace a pre-installed Chrome, and can argue their exclusion from the factory image cost them the distribution that decides browser market share.
• Manufacturers and would-be Android forks that were contractually barred from shipping unapproved versions of the system — the anti-fragmentation restraint — and can point to products that never reached the market as a result.
• App stores and app developers shut out by the Play Store's privileged, mandatory placement, several of whom (Aptoide among them) were complainants and have pursued Google in national courts already.

None of these claims is automatic, and I want to be precise about that, because the gap between "has standing" and "wins" is exactly the gap this ruling does not close. Causation and quantum in a digital-default case are genuinely hard: Google will argue, plausibly, that users chose its search because it was better, not because it was pre-installed, and untangling the two is expert-witness territory that can run for years. What changed on 2 July is not that these plaintiffs will win. It is that they can now start on the near side of the hardest wall, with the unlawfulness already stamped and unchallengeable, and with a limitation clock that EU law is careful to keep open until precisely this moment of finality. The expensive, uncertain part of a competition case just became free and certain for everyone Google harmed.

The enforcement is a museum piece. The precedent is not.

There is a genuine irony in the timing that is worth stating plainly rather than scoring. As a piece of live regulation, the Android decision is almost historical. The conduct ended in 2018; Google long ago unbundled the apps, introduced choice screens, and started charging separately for the Play Store in Europe. And the behaviour the case was about — self-preferencing, mandatory defaults, tying your own services to the platform everyone needs — is now governed going forward by a newer instrument that does not require a decade of litigation to bite: the Digital Markets Act, under which Google is a designated gatekeeper and the same practices are simply prohibited from the outset, on pain of fines running to a tenth of global turnover. In pure enforcement terms, the Court spent 2026 finalising a case about a world the DMA has already replaced.

But a precedent does not expire when the conduct does, and this one arrives at an unhelpful time for the industry it most resembles. The fight the Android case was really about — who gets to be the default on the device in your hand, and what the platform owner may demand in exchange — is not a settled question from the last decade. It is the live question of this one, moved up a layer from search boxes and browsers to AI assistants. The same companies are now wiring the same phones so that one assistant answers when you speak to the device, and are signing the same kind of pre-installation and revenue arrangements to make sure it is theirs. The Court of Justice has just confirmed, at the highest and final level, that using control of a dominant platform to hard-wire your own service into the default position is an abuse that carries liability — and, now, transferable liability that private parties can enforce. That holding does not care whether the service in the default slot is a search engine or a chatbot.

This is the part that outlasts the fine, and it is the part that fits the pattern Brussels has been repeating for a decade. Europe rarely builds the dominant platform. What it builds is the rule about what the platform may do with its dominance, and because no one engineers a separate phone for the European market, the rule tends to set the floor everywhere. A €4.1 billion penalty for conduct that stopped in 2018 is, on its own, a slow and unremarkable outcome. A final, unappealable, privately enforceable finding that platform defaults are a place where dominance becomes liability — delivered in the exact year the industry is re-fighting the default war over AI — is not slow at all. It is early. Google will pay the fine and forget it by the next earnings call. The finding it spent eight years failing to erase is the one its rivals, and its regulators, will be quoting for the next eight.