On September 2, 2026, Judge Leonie Brinkema handed the Department of Justice its third straight loss on the question of whether a Big Tech monopoly should actually be broken up. Google will not have to sell AdX. It will not have to open-source DFP's auction logic. Ad Manager — the exchange and the ad server, still bundled under one roof — stays exactly where it has always been: inside Google (AdExchanger). Instead, the court accepted "most" of the parties' proposed behavioral remedies — a real-time bid feed for rival ad servers, the removal of Unified Pricing Rules, a monitoring trustee — the details of which remain sealed pending a redacted opinion due in the coming weeks (PPC Land).

If you have watched antitrust enforcement against the platforms for the last few years, this outcome should not surprise you. It is the third time in this cycle that a federal court has found a dominant company liable for illegal monopolization and then declined to make it give anything up. Judge Amit Mehta did the same thing with Google Search in 2024, ordering data-sharing instead of a Chrome or Android divestiture (Congress.gov). Brinkema herself was explicit about the logic: breaking up AdX or DFP risked harming the small publishers who depend on them today, any forced sale to a new owner introduced its own set of complications, and behavioral fixes could be implemented years faster than a divestiture could survive appeal. The market got its answer. Ad Manager is not going anywhere.

What the market has not fully absorbed yet is what that answer actually means for the people who buy and sell inside it.

The remedy addresses the crime. It does not address the incentive.

Behavioral remedies are, definitionally, promises about future conduct. Google will make bid data available. Google will stop enforcing pricing rules that constrained publisher-side price discovery. Google will accept a trustee to monitor compliance. These are meaningful, and they are not nothing — PubMatic, which competes directly with AdX, called the court's approach a path toward "a level playing field for all market participants" (AdExchanger). But a promise about conduct is not a change in structure. Google still owns the exchange. Google still owns the ad server that decides which bid wins. Google still owns the buy-side demand — Google Ads and DV360 — that both of those systems are built to serve. The court's own liability finding, entered in April 2025, was that Google illegally monopolized the publisher ad server market, illegally monopolized the ad exchange market for open-web display, and unlawfully tied the two together in violation of the Sherman Act (PPC Land). None of the structural facts that produced that finding have changed. Only the rules governing how the same structure is allowed to behave have changed — and, per multiple accounts of the trial, largely on terms Google itself proposed.

This is precisely the distinction we spend the most time on with clients, and it is worth restating plainly here: commercial bias is not a compliance failure that a monitor can catch and correct. It is a structural feature of a system where the same company runs the auction, prices the inventory, and represents a share of the demand bidding into it. A dominant company being told to behave better inside a structure that still rewards routing value toward itself is not the same as removing the incentive to do so. Even Thomas Höppner, the competition lawyer who has litigated against Google in Europe, put the skepticism plainly: "I'm curious why the judge concluded behavioural remedies alone will be sufficient" (PPC Land). The European Commission reached the opposite conclusion in its own €2.95 billion decision last September, finding that behavioral fixes had historically let Google "modify practices to avoid detection while preserving the effects" (PPC Land). Two regulators, looking at the same company, reached opposite conclusions about whether a promise is enough. That disagreement alone should tell agencies and brands everything they need to know about how much certainty to place in a sealed opinion that has not even been published yet.

The take rate survives. So does the reason to route around it.

Set the legal theory aside for a moment and look at the economics. AdX still charges publishers roughly a fifth of gross revenue for access to Google's demand (Reuters). Ad Manager remains an integrated product, so the take rate, the auction logic, and the buy-side relationship all sit inside a single, self-interested owner. That was true the day before Brinkema's order. It is true the day after. Nothing in a behavioral remedy — no matter how well designed the eventual Final Judgment turns out to be — removes the plain commercial reality that value tends to flow toward the platform that controls the pipes, unless someone with genuine visibility is watching where that flow goes.

That is a live, expensive problem for agencies and brands right now, independent of how this litigation resolves on appeal. Google's own Network revenue — the segment capturing third-party advertising through this exact stack — fell 4% year over year in the first quarter of 2026, attributed to shifting traffic patterns rather than any remedy (PPC Land). The open web is already reallocating itself. The follow-on damages litigation tells the same story from a different angle: OpenX, PubMatic, Magnite, Raptive, Index Exchange, Vox Media and The Atlantic have all filed claims built on Brinkema's liability findings, several alleging billions of diverted impressions and depressed yield (PPC Land). None of that money finds its way back to the agencies and brands who overpaid for a supply path they never chose. It funds a decade of litigation. The remedy for the industry's overpayment problem was never going to arrive in a courtroom. It has to be built.

What we would tell a client to do Monday morning

We are not in the business of predicting appellate outcomes, and we would caution any agency against building a 2027 media plan around the assumption that a sealed opinion, still subject to a Final Judgment process that runs at least through October and an appeal Google has already signaled it intends to file, will meaningfully change how Ad Manager behaves in the next two quarters. What we would tell a client is simpler: the ruling confirms, rather than resolves, exactly the structural bias MCG was built to sit outside of.

MCG operates as a transparent layer nested over the SSPs — not another hop competing for margin inside the chain, and not a walled garden asking a buyer to abandon the DSPs and technology they already run (wermcg.com). We do not ask an agency to rip out The Trade Desk, DV360, or Amazon DSP. We shape and curate the supply feeding whichever platform the buyer already uses, strip structural SSP and DSP bias out of the auction before the bid is placed, and issue standard deal IDs directly into the agency's existing seat — no new contracts, no new technology, no workflow changes. That distinction matters more, not less, in a market where the dominant exchange has just been told by a federal court that self-interested conduct is a behavioral problem to be monitored rather than a structural one to be removed. Someone still has to look at the supply path with no stake in where the money lands, and act on what they see, in real time, on the buyer's behalf.

That is also, not incidentally, where genuine differentiation now lives for the agencies reading this. Very few agencies have a dedicated capability for shaping their own supply chain; most are, structurally, price-takers on whatever an SSP or exchange decides to show them. An agency that can tell a client it has upstream visibility and control over where its media dollars actually travel — independent of Google's compliance calendar — has a structural advantage that a competitor reciting the same platform-provided reporting simply does not. That advantage compounds directly into the agency's economics as well: the market persistently overspends for the quality it procures, and closing that gap creates efficiency an agency can reinvest into performance, pass through as a client-retention tool, or capture as margin. We built the layer to be additive, not disruptive, precisely because agencies do not need one more platform to manage. They need a partner that makes their existing stack smarter and their existing supply cleaner, and that shows up in outcomes their clients can see even if they never know we are there.

A ruling that changes the paperwork, not the physics

Brinkema's decision will matter enormously to Google's legal exposure, to the shape of the eventual Final Judgment due by October 2, and to whatever the European Commission does next with its own, harsher read of the same conduct (PPC Land). It will matter far less to the daily reality of buying and selling media on the open web, because it leaves the physics of the market exactly where they were: one company still owns the exchange, the ad server, and a dominant share of the demand competing inside both. Behavioral remedies are a bet that disclosure and monitoring can substitute for structural separation. Maybe they can, eventually, once the sealed opinion becomes public and a trustee is actually watching. Agencies and brands do not have the luxury of waiting to find out. The dominance the DOJ tried and failed to dismantle in a Virginia courtroom is the same dominance MCG was built to help our clients see around, right now, with the platforms they already have.

Donald Williams is Chief Strategy Officer and Executive Advisor at Media Consulting Group (MCG).