The strangest thing about Google’s antitrust story is this: the court has ruled the company a monopolist, and its Search still holds nearly 90% of the market. Whether the ruling changes anything, and where it ends up, is an unavoidable question for anyone trying to read the risks around this company.
This piece uses a single timeline to separate what’s already settled from what’s still in flux: exactly what the remedy covers, why Gemini got written in, and whether the real hit to Search is as big as the outside world imagines. Up front: this is a plain-language explainer and a relay of the legal proceedings—it doesn’t predict who wins or loses, and it makes no investment calls. If you’d like to get to know Google as a company first, see What kind of company is Google.
The one-line framing: this is a behavioral remedy, not a breakup; its bite is more chronic than acute, and the real verdict still rides on the appeal and those EU fronts—far from over.
The ruling, on a timeline
Lay out the key milestones and the rhythm of the whole thing becomes clear.
| When | What happened |
|---|---|
| 2024/8 | Federal judge Mehta rules that Google unlawfully maintained a monopoly in the general search market |
| 2025/9 | The remedy lands: a ban on exclusive default contracts and a requirement to share search data, but no breakup of Chrome/Android and no mandatory choice screen |
| 2025/12 | The final judgment is formally entered; the six-year remedy-oversight period begins roughly 60 days later (early 2026) |
| 2026/1 | Google files an appeal with the federal circuit court of appeals, challenging the data-sharing and oversight mechanisms |
| 2026/2 | The DOJ and several states file a cross-appeal seeking stronger remedies (including the breakup the judge declined) |
| 2026/5 | Google submits an opening brief running over a hundred pages, arguing that both the liability finding and the remedy should be overturned; oral argument is not yet scheduled |
From the ruling to a final resolution, this road is expected to run into late 2026—or even early 2027—before there’s an appellate outcome.
What the remedy actually covers
A lot of people assume “found to be a monopoly” equals “broken up.” This time it doesn’t.
The federal judge chose a behavioral remedy—changing how Google operates rather than forcing it to sell off assets. The key provisions: Google is barred from signing “exclusive” default-search contracts with the likes of Apple and Mozilla; no default contract may run longer than a year or auto-renew; and Google must share a one-time snapshot of its search index plus rolling user-interaction data (queries, clicks, dwell time) with “qualified competitors”—but not ad data, and competitors may not use that data to train large language models. The court also set up a six-year technical committee to oversee compliance.
Just as important is what the court did not do: it declined to force Google to sell Chrome or Android, and it declined to mandate a “choice screen” that would let users pick their search engine. The reasoning on the latter is pragmatic—Europe has mandated a choice screen since 2020, and Google’s mobile search share has held above 97%. The effect was limited.
Why Gemini got written into the ruling
On its surface this case is about search, yet it pulled AI into the circle too—its most direct point of contact with the AI race.
The court’s reasoning was that “Gemini operates a lot like search,” so it folded Gemini and its successor products into the scope of the remedy, the intent being to stop Google from using its search-market monopoly to squeeze out other AI assistants. Concretely, the court barred Google from the bundling move of “give you search perks, but you have to default exclusively to Gemini.” But note the nuance: Google can still pay to win default placement for Gemini—as long as it isn’t conditioned on exclusivity and the contract doesn’t exceed a year.
In other words, the ruling restricts “trading one monopoly for another,” not Google’s right to promote Gemini. That line is the key thing to watch as Gemini builds out its distribution.
The real impact on Search: not that dramatic
To talk about the impact, you first have to dial down the emotion and look at how the rules are actually built.
The most immediate variable is Apple. Google pays Apple roughly $20 billion a year for default search in Safari; the remedy doesn’t ban that money, but it switches the deal to annual renegotiation with no exclusivity—introducing a round of bidding pressure every year. In the short term, even if users are given a choice, most will probably still pick Google, which is exactly why Europe’s choice screen had limited effect; the real uncertainty sits in the medium term, depending on whether Apple gets serious about backing another search engine or building its own.
As for how big the hit to revenue is, analysts diverge wildly—estimates range from “barely noticeable” to “a double-digit-percent impact on operating profit.” The premises differ, and so the conclusions land worlds apart; there’s no widely accepted bottom line right now. Two facts are firmer. First, Google’s global search share remains high (although some survey-based studies show a decline, survey methodology differs from actual traffic measurement, so those numbers need to be read separately). Second, the bite of this remedy is more chronic than acute—it works by eroding the moat slowly through year-by-year renegotiation and data sharing, not by one decisive cut.
Two more fronts
The search case is only part of Google’s regulatory pressure; two other threads bear on AI just as much.
One is the adtech case. In April 2025, in a separate matter, another federal judge ruled that Google holds a monopoly in the publisher ad-server and ad-exchange markets and had unlawfully tied the two together; the DOJ has asked it to sell off the ad exchange, but as of now the remedy ruling still hasn’t come down.
The other is in Europe. Under the Digital Markets Act (DMA), the EU opened two proceedings against Google in early 2026: one requiring Android to grant third-party AI services equal software and hardware interoperability (wake words, on-screen content awareness, becoming the default assistant, and so on)—the documents name OpenAI and Anthropic directly as beneficiaries; the other requiring Google to share data, on fair terms, with rival search engines and search-capable AI chatbots. The EU expects to reach a final decision by mid-2026.
What hasn’t been finally decided
The discipline this topic demands most is to keep “settled” and “unsettled” apart:
- The search-case appeal: both Google and the DOJ have appealed, oral argument is not yet scheduled, and an outcome is expected at the earliest in late 2026 to early 2027.
- The adtech-case remedy: the judge has reservations about a structural breakup (hard to find a buyer, hard to enforce while on appeal), the ruling keeps getting pushed back, and it still hasn’t been published.
- The EU DMA final decision: a preliminary conclusion is out, but the final wording isn’t locked in.
- Analysts’ revenue-impact estimates: the spread across firms is enormous and the premises vary, so no single number should be treated as the bottom line.
These are all still in transit, and any assertion like “Google is about to be broken up” or “the ruling has no impact at all” deserves a question mark first.
小企鵝’s take
Pull this case back far enough and what it’s really testing is an old question: can a behavioral remedy actually rein in a winner?
Google’s position is peculiar: it’s been whistled and warned by the referee, yet it’s still ahead on the field—handling the litigation with one hand while sprinting at full speed on AI with the other. Folding Gemini into the remedy effectively sets the rules for the next battlefield (AI assistants) ahead of time, and that’s the most forward-looking stroke of this ruling. But behavioral remedies have always been slow to bite—Europe’s choice screen is the cautionary precedent. The real test will land around 2027, when three things come due in sequence: the appeal becoming final, the Apple contract getting renegotiated, and the EU dropping its hammer. Until then, the score in this fight doesn’t count yet.
Further reading: What kind of company is Google, Where the $190 billion in capex goes, Gemini vs. ChatGPT: who has more users.