The Google Antitrust Cases Explained: Search, Android, AdTech and What Comes Next

The Google antitrust cases span three continents, two decades and every major theory of platform abuse: self-preferencing, tying, exclusive defaults, and vertical integration of an entire ad market. This guide organises the sprawl into one map — what each case alleged, what was decided, what was annulled, and what actually changed — with the EU, US and Turkish decisions as anchors. It is the second article in the landmark-cases pillar of our Competition & Antitrust hub.

What did the EU Google Shopping case decide?

The 2017 Shopping decision — €2.42 billion, then the largest abuse fine ever — held that Google demoted rival comparison-shopping services in its general results while displaying its own Shopping units prominently, leveraging search dominance into an adjacent market. The theory now has a name every platform knows: self-preferencing.

The Court of Justice definitively upheld the decision in September 2024, entrenching self-preferencing as an abuse category in EU law without requiring proof that Google’s conduct met the stricter refusal-to-deal test. The remedy — an auction in which rivals (and Google itself) bid for Shopping slots — remains the case’s weak flank: comparison sites still complain it changed little, an outcome that became Exhibit A for the argument that conduct remedies in digital markets arrive too late and bite too softly, and thus for the Digital Markets Act.

What happened in the Android and AdSense cases?

Android (2018, €4.34 billion — trimmed to €4.125 billion by the General Court in 2022) condemned the contractual architecture of mobile dominance: requiring Google Search and Chrome pre-installation as the price of the Play Store, paying for exclusivity, and “anti-fragmentation” agreements blocking OEMs from selling forked-Android devices. The echo of Microsoft’s OEM contracts is deliberate and acknowledged.

AdSense (2019, €1.49 billion) attacked exclusivity clauses that stopped publishers using rival search-ad intermediaries — but the General Court annulled the fine in September 2024, finding the Commission had failed to prove the clauses’ effects across their full duration. The annulment matters beyond Google: it shows EU courts genuinely policing the Commission’s economic evidence, and it is the standing reminder that headline fines are provisional until the last appeal — a caution that applies equally to the €2.95 billion adtech decision of September 2025, which Google is contesting.

What did the US search case decide — and what are the remedies?

In August 2024, Judge Amit Mehta ruled that Google violated Section 2 by maintaining its general-search monopoly through exclusive default agreements — above all the multibillion-dollar arrangement making Google the default on Safari and Android devices. “Google is a monopolist, and it has acted as one to maintain its monopoly” became the era’s defining sentence.

The September 2025 remedies order chose conduct over structure: no Chrome divestiture, no Android spin-off. Google must instead share search index and interaction data with qualified competitors, syndicate results, and abandon exclusive default contracts — defaults may be bought, but not exclusively and not with retaliation. The court’s reasoning leaned on generative AI: with ChatGPT-class assistants already eroding search’s gatekeeping position, structural surgery looked riskier and less necessary than opening the data moat. Appeals will run for years; but as in Microsoft, the liability architecture — defaults as exclusion, scale as the barrier — is already being cited across the field.

What is the adtech case — and why is it the structural one?

Google’s advertising-technology stack integrates every layer of the open-web display market: the dominant publisher ad server (DFP), the dominant exchange (AdX), and the dominant buy-side tools. Both the DOJ and the European Commission concluded this vertical integration was abused — self-preferencing AdX in DFP’s auctions, tying access, and manipulating auction mechanics against rivals and publishers.

Judge Leonie Brinkema found Section 2 liability on the publisher-side markets in April 2025, and the DOJ pressed for divestiture of the sell-side stack in the remedies phase. The Commission’s September 2025 decision fined Google €2.95 billion and — unusually — stated openly that behavioural fixes may not suffice, keeping divestiture on the table. Adtech is thus the live test of the question Shopping raised and Microsoft first asked: whether conflicts of interest embedded in vertical integration can be remedied by rules of conduct at all, or only by separation — the subject of our breakups article.

How do the Google cases connect to the DMA?

Causally. Each litigated theory — self-preferencing, pre-installation tying, exclusive defaults, data leverage — reappears in the Digital Markets Act as a per-se obligation on designated gatekeepers, enforceable without proving dominance, effects or duration. The DMA is, in substance, the Google case law converted from judgments into a rulebook.

Google’s DMA compliance is itself contested — its search-results changes and Play Store steering rules drew the Commission’s early scrutiny, and a non-compliance finding carries fines up to 10% of worldwide turnover. The practical state of play for gatekeepers, specification decisions and the first DMA fines is covered in our DMA enforcement tracker; the doctrinal foundations sit in the abuse-of-dominance pillar.

What should businesses take from the Google saga?

Three transferable lessons. Vertical integration plus gatekeeping equals a target: operating the marketplace and competing in it — the structure at issue in Shopping and adtech — is now the most litigated configuration in competition law, whatever your industry. Defaults and pre-installation are market power: any firm paying for or contracting exclusivity over a distribution chokepoint should model the search case before signing.

And remedy design is where cases are really decided: Google has lost liability fights repeatedly while containing structural damage — auctions, choice screens, data sharing instead of breakups. Whether that containment survives the adtech endgame is the most consequential open question in global antitrust; the answer will calibrate every future platform case, and we track the candidates for structural relief in our breakups analysis.

How do the Google cases define digital market boundaries?

Narrowly, and that has been decisive. In the US search case the court accepted general search services as a market — excluding Amazon product searches and social discovery — and general search text ads beside it. The EU adtech decisions carve publisher ad servers, ad exchanges and buy-side tools into separate markets, each dominated individually. Once markets are drawn at the layer level, shares of 80-90% follow, and dominance is nearly automatic.

Google’s standing counterargument — competition is one click away, and the true market is attention or advertising at large — has now lost in Brussels, Washington and Ankara alike. The doctrinal settlement matters beyond Google: layer-by-layer market definition is how every integrated digital stack will be analysed, which is why vertically integrated firms increasingly face dominance findings in intermediate markets end-users never see. The tools courts used are unpacked in our market-definition guide.

Is AI Google’s defence — or the next case?

Both, simultaneously. The 2025 search remedies opinion leaned on generative AI to reject a breakup: if ChatGPT-class assistants genuinely erode the search gatekeeping position, structural surgery risks solving a receding problem. Google’s counsel argued the point; the court partly accepted it; and remedy design (data sharing, no exclusive defaults) was calibrated to help AI-era entrants contest defaults.

Yet the same technology is generating the next docket: publisher complaints over AI Overviews repurposing content without traffic, scrutiny of default placements for AI assistants on devices, and competition-authority studies of foundation-model concentration in compute, data and talent. Regulators explicitly frame the moment as a do-over — having watched search, social and mobile tip into monopoly, they intend to police AI’s chokepoints earlier. That agenda, from cloud partnerships to chip supply, is the subject of our AI and competition law analysis.

What should advertisers and publishers do while the cases run?

Treat the decisions as commercial leverage, not spectator sport. Publishers should quantify their exposure to the condemned adtech mechanics (auction changes, fee opacity) — that analysis grounds both damages claims, several of which face limitation deadlines tied to decision dates, and renegotiation of ad-stack terms. Advertisers should benchmark their buy-side fees against the trial record’s disclosed take rates, the rare moment such data is public.

Both should scenario-plan an adtech divestiture: contract assignment, data portability and auction-dynamics changes would arrive on a court timetable, not a commercial one. And any business relying on Google defaults — browsers, OEMs, carriers — should model the post-exclusivity bidding landscape the US remedies create, where default slots become genuinely contestable auctions for the first time in twenty years.

Frequently Asked Questions