In the second half of September, the first thing circulating among media buyers was scary, that Google's ad business had been broken up. By the time the full memorandum became public around September 16, the conclusion turned out to be the opposite, there is no breakup. The ad tech antitrust case the Justice Department filed in 2023 and won on liability in April 2025 has reached a key step in its remedies phase, and the direct impact of that step on ordinary ad accounts is much smaller than the headlines suggested.

I went through the full 106-page memorandum, Reuters' coverage, and several trade outlets' write-ups. This piece is not about who won the legal argument. It answers three questions, what changed, what did not, and what export advertisers should do now.

What the Memo Says: No Breakup, Behavioral Fixes and a Monitor

Start with the conclusion. The judge rejected every structural remedy, the sale of AdX, the open-sourcing of DFP's final auction logic, and the contingent divestiture that would have followed. In each case, no. The memo's reasoning is that a forced breakup is neither realistic nor needed, and that the behavioral remedies, as modified by the court, are sufficient to restore competition to the markets that were harmed. The outcome surprised many, because it means AdX and DFP stay inside the same company, and the fee structure does not move.

The behavioral remedies that were adopted fall into three broad categories and five groups of provisions. These five groups are the part that matters.

The first group is prohibitions. Google may no longer enforce the policies or contract terms that tied DFP to AdX, may not reinstate First Look or Last Look, and must deprecate Unified Pricing Rules (UPR) and never reintroduce them. For publishers this is not new, UPR was already removed at the end of 2025; the judgment makes non-reintroduction binding.

The second group is interoperability. AdX must interoperate with rival publisher ad servers and with header bidding mechanisms such as Prebid, and DFP must interoperate with Prebid as well. On timing, Google committed to completing the DFP-to-Prebid part within 12 to 15 months, and the court accepted that shorter timeline.

The third group is data sharing. The publisher data-sharing obligation proceeds on Google's narrower version, and DFP's decision documentation must be made transparent, with the format and disclosure frequency to be worked out by the parties and subject to the technical committee.

The fourth group is non-discrimination. Data signals must be used without regard to whether a counterparty uses Google's tools; AdX must pass bids to rival servers and header bidding without varying revenue share, latency, or response frequency; and DFP must serve ads according to publishers' preferences regardless of whose bid wins.

The fifth group is buy-side constraints, which apply only to the Google Ads demand-side system, called AdWords in the memo. DV360 is out of scope. The system may not build direct bidding integrations with DFP, must bid and use first-party data on a non-discriminatory basis, and may not favor its own routes.

The oversight mechanism is a monitor plus a three-member technical committee, with each side selecting one member and those two jointly selecting the third. The monitor files a work plan within 45 days, reports in writing to both sides quarterly, and must notify plaintiffs immediately upon reason to believe Google is not complying. The term is six years, against the plaintiffs' request for 15, and the court can extend it if the judgment is not fully satisfied. The judgment takes effect 60 days after entry, but the obligation to begin forming the monitor and committee starts effectively immediately.

One boundary matters. All of the above is what the memo has disclosed so far. The final judgment text still has to be jointly submitted within 30 days, unresolved provisions will run as competing versions labeled by proponent, and the media's stated deadline is October 2. In other words, the result is settled while the details are still landing.

Four Practical Impacts for Advertisers

First, the supply and auction environment. The substance of the interoperability provisions is that AdX's real-time bids must flow equally to third-party ad servers and header bidding. For teams doing supply-path optimization, the way open-web inventory is organized may gain options over the medium term, and the value of integrating third-party exchanges and header bidding will shift slowly. In the short term nothing moves, coverage describes the supply path as identical to the day before, the take rate is unchanged, and the industry estimate for opening AdX to Prebid runs in years.

Second, data and measurement. The data-sharing obligation, the non-discrimination rules for data signals, and the DFP documentation transparency all point the same way, toward gradually higher visibility into how auction decisions are made, and possibly more room for third-party monitoring and verification. Note that DV360 sits outside the buy-side constraints, so your Google buying tools do not change immediately because of this ruling, and your measurement stack does not need to move.

Third, fees and terms. Rates are unchanged and the ruling requires no adjustment. The impact is in cadence. For the next six years, changes to terms and products around this ad stack will sit under the monitoring mechanism, making shifts easier for outsiders to see, though when and how they arrive depends on the final judgment text and its execution.

Fourth, planning premises. Over the past two years, some teams treated a possible Google breakup as a planning premise, bending vendor evaluations, media architecture, and contract terms around it. That premise is now removed and the related assumptions need updating. The opposite reading fails too, regulation landing does not mean near-term rules chaos; the path is behavioral remedies plus six years of monitoring, executed in years.

Three Narratives to Put Down

Narrative one, Google was broken up. It was not. All three structural proposals were rejected, and neither AdX nor DFP moves.

Narrative two, the breakup is merely postponed and will happen eventually. There is no basis for this. The structural question has been decided in this case; what proceeds next is the final judgment text and the appeal, not another breakup process waiting in line.

Narrative three, regulation landing means the rules of buying change immediately. The direction is reversed. The changes concentrate in the supply and demand mechanics of the open web, and they execute over years. Account-level and self-serve surfaces do not change in the short term.

Configuration Advice for the Uncertain Window

On dependency, evaluate your reliance on any single platform with ordinary commercial logic, and do not add or subtract because of this case. Teams doing supply-path optimization can simply put AdX-to-third-party-server and AdX-to-Prebid interoperability on next year's watchlist.

On budgets, make no preventive migrations. Neither the court nor the coverage establishes a basis for price or efficiency changes, and we make no share or price predictions. Until a basis exists, current allocations do not need to move because of the news.

On measurement, keep your stack as is, and treat the monitoring channel as a new source of intelligence. Quarterly reports and the DFP documentation are things the market did not have before.

On contracts, if a vendor agreement or agency contract defines a trigger around platform status changes, review whether its definition covers this scenario.

Action Lists for Three Roles

In-house brand teams. First, update the channel assumptions document and remove the breakup expectation. Second, keep account structures and bid automation running as usual; the only thing to watch is how the buy-side non-discrimination and routing provisions later translate into product details, especially for automated rules built on Google signals. Third, add the case's key dates to the quarterly risk list.

Agencies. First, give clients three lines, no breakup, behavioral fixes plus a monitor, no short-term operating change. Second, do not write any breakup-upside scenarios into media plans, and do not promise price shifts. Third, track the early-October final judgment text and how the six-year monitoring runs, updating internal knowledge when reporting definitions are affected.

Independent sites and publisher-side teams. First, interoperability affects you most directly, AdX connecting equally to third-party servers and header bidding, plus the UPR removal, widens the room for ad server and floor strategy, so put multi-stack options on the shortlist. Second, Prebid integration progress and the data-sharing terms are the two concrete things to track. Third, do not rebuild systems during the transition; wait for the final judgment text.

Timeline: Execution, Appeals and Ongoing Proceedings

On September 2, the court issued its order, rejecting structural remedies and adopting most behavioral ones, with the memorandum sealed for 14 days. On September 16, the 106-page memo was unsealed, detailed coverage began, and the case moved from outcomes to details.

Within the next 30 days, the parties must jointly submit a final judgment text, with the media's stated deadline of October 2. It takes effect 60 days after entry, the monitor formation obligation starts immediately, and the monitoring term runs six years with possible extension.

Two parallel tracks continue. One is the appeal, Google has said it will appeal the liability findings, which will shape how the six-year monitoring period plays out procedurally. The other is the EU's parallel case, where proceedings continue in Brussels; some coverage reads the structural hope as shifting there, which is a media view, and the two proceedings do not bind each other.