Judge Leonie Brinkema has answered the question that has hung over the Google advertising-technology case since April: if the government cannot break up the business, what does a remedy actually look like? The 106-page decision unsealed Wednesday in Alexandria, Virginia, says it looks like six years of behavioral rules, a technical committee, and an antitrust compliance monitor, all paid for by the company being monitored.
The ruling completes the liability-and-remedy arc of the case. Brinkema found in April 2025 that Google unlawfully monopolized the publisher ad-server and ad-exchange markets. This month she rejected the Justice Department's demand that Google sell its ad exchange, AdX, and now the full opinion fills in what she ordered instead. Our earlier piece covered the no-divestiture headline; the unsealed decision shows the remedy's real engine is the oversight apparatus around the behavioral rules.
The rules themselves
The ordered changes target the choke points the liability ruling identified. Google must make its ad tools work more closely with rival products and share more data with publishers. It can no longer require a website that uses its ad server to also use AdX, breaking the tie that locked publishers into the full Google stack. Rival ad servers must be able to receive AdX's real-time bids, and Prebid, the open-source header-bidding tool, must be able to solicit real-time bids from AdX directly. Restrictions land on some AdWords buying tools, while the DV360 demand-side platform escapes curbs.
The mechanism language matters more than the list. Brinkema wrote that the behavioral remedies "will be sufficient to effectively pry open to competition" the markets harmed by Google's conduct. A judge choosing behavioral remedies over divestiture is not unusual; a judge writing that they will pry open a market is making a prediction, and predictions about market structure are exactly what the next six years will test.
Why no sale of AdX
The government had argued that AdX, where publishers pay a 20 percent fee on transactions, should be sold outright. Brinkema said divestiture was "neither realistic nor needed." The government, she found, failed to identify a viable buyer, and a sale would take too long to matter in a market being reshaped by AI-driven disruption.
That reasoning is consequential beyond this case. The Justice Department's remedy demands in the big platform cases, and the search case in particular, have rested on structural change: sell assets, restore competition by breaking the bundle. Brinkema's answer, in the case that proceeded fastest to remedy, is that structure is sometimes slower and less certain than a sufficiently sharp set of rules. If that judgment holds up on appeal, it will shape how the other monopolization cases are argued.
The monitor, and who pays for it
The decision requires an internal antitrust compliance officer and a technical committee, plus the external monitor, all at Google's expense. The monitor's role is the part Brinkema trimmed: the government sought expansive oversight authority, and the court ordered a more limited mandate, reasoning that the gravity of the violations warranted monitoring without handing the monitor the keys to the business.
Company-funded monitors are a familiar instrument with a known weakness. The monitor reports to the court, but the company pays the invoices and staffs the compliance office that feeds the monitor information. The arrangement works when the court stays engaged and the monitor's reports become public records of progress or failure. It fails quietly when compliance reporting becomes a quarterly ritual nobody reads against the underlying market data.
The six-year term sharpens that risk. The Justice Department and the state plaintiffs asked for a 15-year decree. Brinkema set six. The rules apply globally, not just in the United States, which is broader than many expected. But six years is one market cycle, maybe two, in a business that is being reshaped by AI while the decree runs.
What each side says now
The Justice Department called the decision a "significant victory," which is what a government calls a case where it got liability, a monitor, and interoperability rules but not the sale it asked for. Google said it disagrees with the liability ruling on its Google Ad Manager tool and will appeal, which keeps the core question open: the remedies presume the liability finding, and an appellate reversal of liability would take the remedy architecture down with it.
Both sides have 30 days to file a proposed final judgment. The details that follow, the exact scope of the monitor's access, the technical committee's composition, the reporting schedule, are where decrees of this kind are won and lost in practice. The order says Google must interoperate. The proposed judgment will say what happens when it does not, and who checks.
A quick map of the market being pried open
The ad-tech pipeline the case concerns is the open display market: the auctions that decide which banner and video ads appear on news sites and apps in the instant a page loads. A publisher uses an ad server to decide which ads to show and logs what runs. Exchanges run the auctions that sell the impressions. Demand-side tools let advertisers bid into those auctions. Google operates the largest player in every layer, which is what made the tying allegations consequential: a publisher on Google's ad server was steered toward Google's exchange, and Google's exchange took its 20 percent cut from auctions rivals could not meaningfully reach.
The interoperability orders attack the steering, not the ownership. A rival ad server that can query AdX's real-time bids becomes viable for a publisher who wants Google's demand without Google's stack. A publisher who can use Prebid to solicit AdX bids keeps the open-source auction infrastructure while still selling through Google's exchange. The orders do not require Google to lose the ad revenue. They require the auction to admit competitors at the points where the market was closed.
The monitor sits on top of this because the rules are technical in a way that invites technical evasion. Auction latency, bid-sharing formats, fee structures, the order in which requests are routed, all of these can be adjusted by engineers in ways that make a formally compliant system behave like the old closed one. The decree's bet is that a monitor, a technical committee, and six years of court jurisdiction can keep the engineering honest. Whether that bet pays is what the next six years will show.
The precedent the case now sets
The decision lands at a moment when the government's structural-remedy playbook is under active debate. The Justice Department has sought breakups or divestitures in its major platform cases, arguing that behavioral rules have failed before. Brinkema's answer in the first big ad-tech case to reach remedy is a controlled experiment in the opposite direction: a bounded set of behavioral rules, a monitor with a limited mandate, and a six-year clock.
The comparison to the Microsoft decree of two decades ago is unavoidable, and the comparison cuts both ways. That decree imposed behavioral rules and a monitoring apparatus, and the market eventually shifted anyway, with browsers and mobile changing the battlefield the decree did not anticipate. Behavioral remedies have a habit of being overtaken by the market they regulate. Brinkema's opinion acknowledges the AI disruption racing through advertising right now, and it is precisely that disruption that makes a six-year decree feel simultaneously long and short: long enough to outlast the litigation, possibly too short to matter, and almost certainly outlasted by the technology.
The appeal keeps the outcome unsettled, but the architecture of the remedy is now public, argued, and written in a judge's own words. Whatever the appellate courts do, the next monopolization case to reach a remedy will be argued against this one. The monitor, the technical committee, the six-year term, the refusal to sell AdX, all of it is now the baseline from which the government will ask for more and the companies will argue for less. The decree is the precedent, and the monitor is the experiment.
European regulators are watching from a parallel track, and the decree will interact with their proceedings whether or not anyone coordinates it. The European Union's digital-markets rules already impose interoperability obligations on dominant platforms, and the UK competition authority has run its own advertising-market inquiries. Google's ad-tech compliance obligations will now be read against American, European, and British standards simultaneously, each with its own definitions of what a fair auction looks like. A rule designed for Alexandria, Virginia will be implemented in Dublin, London, and Mountain View, and the monitor will not be the only party checking.
The market, meanwhile, is not waiting for the appeal. Publishers have spent years on the wrong side of the 20 percent AdX fee with few alternatives. Rival exchanges and server tools now have a court-ordered entry point into Google's real-time bids. Whether they convert that opening into competition depends on whether the interoperability rules work technically, and whether publishers trust that using them will not come with retaliation they cannot see. A monitor with a limited mandate, paid by the monitored company, is the only thing standing between those rules and their enforcement. The decree is now the monitor.
Primary sources
- Courthouse News for the unsealed 106-page decision, the monitor order, and the scope of the remedies.
- StockAnalysis for the unsealed remedy documents and the behavioral requirements.
- Indian Express for the ruling's terms and the parties' reactions.