The remedy phase of United States v. Google LLC is unfolding in Judge Leonie Brinkema's Alexandria courtroom, and the choice she makes will shape US digital antitrust policy for years. After her April 2025 liability ruling — which found that Google unlawfully monopolized the publisher ad server and ad exchange markets, while clearing it on the advertiser ad network claim — the Department of Justice has pressed for structural divestiture of core pieces of Google's ad tech stack, including the AdX exchange. Google counters that targeted behavioral remedies can address the specific harms the court identified without the cost, complexity, and collateral damage of a corporate breakup.
This is not a question of whether Google's conduct deserves a remedy. The court found that it did. The harder question — the one that should keep both regulators and a pro-innovation policy community up at night — is which remedy actually restores competition in 2026 without freezing the ad tech market in amber or producing years of operational chaos for the publishers and advertisers the case was meant to protect.
What the Court Actually Found
Brinkema's liability opinion was narrower than some of the headlines suggested. She ruled that Google had monopolized two specific intermediation markets — the tools publishers use to manage their ad inventory (the ad server, known as DoubleClick for Publishers or Google Ad Manager) and the exchange where that inventory clears (AdX). She declined to find liability in the advertiser-side network market. That distinction matters: it suggests the court sees the competitive problem as one of plumbing and tying between two integrated tools, not as a sweeping condemnation of Google's entire advertising business.
The DOJ's structural proposal, by contrast, has been broader. Reports indicate the government wants Google to divest AdX at a minimum, and potentially significant parts of its publisher ad server business, to create a standalone competitor capable of breaking the integration the court found anticompetitive. Google has reportedly proposed conduct-based alternatives, including unwinding particular contractual or technical links between its tools.
The Case for Proportionality
Structural divestitures are the heavy artillery of antitrust. They make sense when a market cannot be repaired by rules alone — when integration itself is the harm, and only separating the pieces will restore rivalry. The classic precedents, from Standard Oil to AT&T, fit that mould. The 2001 Microsoft case is the more cautionary tale: the D.C. Circuit reversed the proposed breakup precisely because the remedy outran the proven harm, and the eventual conduct decree worked well enough that Microsoft remained one company while the browser market opened up.
Several features of ad tech argue for caution before reaching for a 1911-style breakup:
- Technical entanglement. AdX, AdSense, DFP and Google's bidding logic share infrastructure, identity systems, fraud detection, and machine learning pipelines built over fifteen years. Separating them is not like spinning off a subsidiary; it is more like extracting an organ.
- A market that is already moving. Apple's ATT framework, Chrome's evolving Privacy Sandbox, the rise of retail media networks (Amazon, Walmart, Uber), connected TV ad stacks, and a much more aggressive supply-side from The Trade Desk and others have reshaped competitive conditions since the conduct at issue began.
- Publisher dependency. Many small and mid-sized publishers rely on Google's integrated stack to monetize at scale. A botched divestiture risks short-term revenue collapse for exactly the constituency the case is meant to protect.
- Appeal risk. An aggressive structural order, on a record built around two specific markets, invites the same kind of D.C. Circuit-style narrowing the Microsoft remedy received — potentially leaving publishers with neither relief nor closure.
What a Proportionate Remedy Looks Like
A pro-innovation, evidence-based remedy would start with the harm the court actually identified and work outward. That means rules that genuinely sever the tying behaviour: open, non-discriminatory access to AdX for rival ad servers; bans on the contract and auction-design tactics the court found exclusionary (including the conduct around "last look" and unified pricing rules); real-time, audited data on auction mechanics; and a monitoring trustee with technical expertise — not just lawyers — to police compliance.
Structural relief should remain on the table as a backstop. If, after a defined period, behavioral remedies fail to produce measurable entry or share gains for rival exchanges and ad servers, the court can revisit divestiture with a fuller record. That sequencing respects both the proven harm and the genuine uncertainty about how to fix a market that is mid-transformation.
Why the Stakes Go Beyond Google
Whatever Brinkema orders will be cited in every major US tech antitrust case for the next decade — from the separate US v. Google search remedy proceedings, to the FTC's Meta case, to whatever comes next for Amazon and Apple. A structural order on a narrow liability record would lower the bar for breakups across the sector. A well-designed conduct order, by contrast, would model a more surgical approach: address the specific exclusionary plumbing, preserve the integrated services consumers and publishers actually use, and let evolving markets do part of the work.
The court has correctly held Google accountable. The remaining task is to ensure the cure is calibrated to the disease. Antitrust is meant to restore competition, not to perform it.