The most consequential antitrust remedy decision of the year ended with a two-page order that did the opposite of what the government asked. Judge Leonie Brinkema, who ruled in April 2025 that Google illegally monopolized the markets for publisher ad servers and ad exchanges, declined to force a sale of either business. She adopted conduct remedies instead, and her stated reasons amount to a concession that the structural fix regulators once reached for no longer fits the market it was designed to correct.
The ruling, issued September 2 in the Eastern District of Virginia, rejected the Department of Justice's demand that Google sell its ad exchange, AdX, and imposed instead what the judge called "most of the parties' proposed behavioral remedies, as modified by this Court." The full opinion is sealed, expected to become public around September 16 absent redactions, and the parties must file a joint proposed final judgment by October 2.
Why the breakup lost
The judge's logic, as reconstructed from her two-page order and the parties' filings, had three parts. First, there was no credible buyer. The most obvious purchaser of a Google ad exchange, Microsoft, would itself trigger an antitrust review, and no other candidate could absorb a business of that scale cleanly. Second, the market has moved. The advertising technology stack at the center of the 2023 complaint already faces structural change from streaming, in-app advertising, and AI-driven buying tools that barely existed when the case was filed. Third, and most practically, a divestiture would be stayed during years of appeals, while conduct fixes could begin to bite within a year.
That last point is the important one. Divestiture is the remedy that is supposed to fix monopoly structure permanently, and conduct remedies are the ones antitrust enforcers distrust because they require ongoing supervision. Brinkema chose the supervised fix on purpose: a slow structural cure was worse than a fast behavioral one.
What the conduct rules reportedly do
Because the memorandum opinion is sealed, the precise obligations are not yet public. Reporting on the case, including at Digiday, describes three categories of limits that track the liability findings. One, restrictions on self-preferencing in ad auctions, meaning Google cannot systematically steer transactions toward its own exchange. Two, obligations to share real-time bidding data with rivals and publishers on terms closer to what Google's own tools enjoy. Three, non-discrimination requirements for rival exchanges and ad servers, including independent auction tools between Google's ad server and its exchange, and the removal of mechanisms like unified pricing rules and first-look or last-look advantages.
The scope fight is already visible in the public record. Google's proposed remedies covered standard banner display advertising only, excluding video, streaming television, and in-app ads. The DOJ wanted every format included. Which scope survived into the final order is among the details the unsealing will reveal, and the dispute previews the compliance fights to come.
The liability underneath the remedy
The remedy rests on findings from April 2025. After a bench trial that ran from September 9 to October 4, 2024, Brinkema held that Google illegally monopolized the publisher ad-server market and the ad-exchange market, and that tying its ad server to its exchange was unlawful per se. The court rejected the government's separate claim over the advertiser ad network. The underlying acquisitions, DoubleClick in 2008 and AdMeld in 2011, were themselves found not to have been illegal.
The numbers from the case explain why the government asked for a breakup. Plaintiffs' evidence put Google's share of the publisher ad-server market at roughly 91 percent from 2018 through 2022, with Google's own figures only slightly lower. AdX's share of the exchange market ran between 54 and 65 percent, roughly nine times its nearest rival. Google's exchange took a 20 percent cut of transactions for a decade, about double the rate of competing exchanges. And yet the ad business in dispute is small for Google itself: Ad Manager accounted for about 4.1 percent of revenue and 1.5 percent of operating profit in 2020, the most recent unredacted figures.
That asymmetry, a business that is modest for its owner and essential to its market, is the classic profile of a market that antitrust law exists to police. It is also why the remedy choice matters more than the liability findings did.
A third breakup denial for the government
The outcome extends a pattern. This is the second time in roughly a year that a federal judge has rejected a DOJ request to break up a Google business, after Judge Amit Mehta declined to force a sale of Chrome in the search case last September. It is the third straight loss for structural remedies against a big technology platform. The European Commission, which fined Google about 2.95 billion euros over related advertising conduct, is now alone among major enforcers in pushing for divestiture.
Reactions split along the lines the litigation established. Google's vice president for regulatory affairs, Lee-Anne Mulholland, said the company was "very pleased the court rejected the DOJ's proposal to break apart tools" that publishers and advertisers rely on. The DOJ said it was "pleased that the court ordered substantial relief" and is evaluating next steps, which may include appealing the remedy's scope. Critics of the outcome were blunter. Sacha Haworth of the Tech Oversight Project said it "takes an Olympic level of mental gymnastics to find Google is operating an illegal monopoly" while leaving its businesses intact.
The compliance era begins
What changes now is procedural and fast. The joint proposed final judgment is due by October 2. The opinion unseals in mid-September. Google has signaled it will contest clauses of the order, and a monitoring dispute over compliance is widely expected, the kind of running fight that defines conduct remedies and the reason enforcers traditionally disliked them.
But the alternative was a divestiture that would spend years on appeal while the advertising market reorganizes around AI-driven buying and selling that the 2023 complaint never contemplated. Brinkema's order is best read as a bet: that a fast, supervised set of auction rules could restore what the monopoly took, and that waiting for a slow structural cure would have meant fixing a market that no longer exists. Whether conduct remedies can police an auction system this complex is the question her sealed opinion will now have to answer over months of compliance, one ruling at a time.
Primary sources
- CNBC for the September 2 order, its two-page form, and the October 2 deadline for the proposed final judgment.
- Digiday for the judge's stated reasons and the reported shape of the conduct remedies.
- Reuters for the May 2025 DOJ breakup proposal and the trial posture.
- IPWatchdog for the April 17, 2025 liability findings and market-share evidence.
- AP reporting, syndicated by ABC27, for the order's public terms.
- The Next Web for the sequence of federal breakup denials and the European Commission's position.