The Department of Justice’s antitrust trial against Google’s advertising technology business begins September 9, 2024. After years of investigation, filings, and pre-trial motions that have produced a remarkable documentary record, the case goes live in the Eastern District of Virginia with Judge Leonie Brinkema presiding.
The trial is consequential for every publisher, DSP, SSP, and advertiser that operates in the open programmatic ecosystem. Understanding what the DOJ is actually alleging — and what the documentary evidence already shows — is essential context for anyone with business exposure to these dynamics.
The Three Core Claims
The DOJ’s case against Google centers on three distinct but related allegations, each supported by internal Google documents that emerged during discovery.
Project Jedi and Last Look: The DOJ alleges that Google’s AdX exchange gave advertisers — specifically Google Ads — a “last look” advantage in publisher auctions. Last look meant that when a publisher’s header bidding stack produced a winning bid from a non-Google DSP, Google’s AdX could see that bid and beat it by a penny, effectively winning the auction with prior knowledge of the competitive bid price. Internal documents labeled this practice “Project Jedi.” The DOJ argues it constitutes market manipulation that disadvantaged non-Google DSPs and generated artificially inflated Google auction wins.
Project Poirot: The DOJ alleges that Google deliberately throttled header bidding — the architectural workaround publishers adopted to route around Google’s AdX control — through a system called Project Poirot. The allegation is that Google Ads reduced bid values on third-party exchanges relative to AdX, effectively steering spend toward Google’s own exchange by pricing non-Google supply below its actual value. Internal documents suggest this was an intentional product decision rather than incidental algorithm behavior.
AdX/DFP Tying: The DOJ’s third major claim is that Google illegally tied AdX (its ad exchange) to DFP/Google Ad Manager (its publisher ad server), requiring publishers who used DFP to effectively transact through AdX. Because DFP dominates the publisher ad server market — it’s estimated to power 90%+ of major publisher ad serving — the tie meant that publishers had extremely limited practical ability to avoid routing transactions through AdX.
What Pre-Trial Disclosures Have Established
The pre-trial discovery process has been unusually productive for public disclosure. Exhibits, depositions, and motions have made portions of Google’s internal communications available — and some of what those documents show is striking.
Internal Google emails cited in pre-trial filings discuss the last look advantage in explicit terms, with Google employees describing it as a mechanism that “could be used to win auctions we otherwise would have lost.” Other documents address Project Poirot as a deliberate competitive strategy to defend AdX market share against header bidding’s encroachment.
The DOJ’s exhibit list is extensive. The internal communications are not in dispute as authentic — Google’s defense is not that these discussions didn’t happen, but that the practices described were either not anticompetitive as a matter of law, or that any competitive advantage was offset by legitimate business benefits to publishers and advertisers.
Google’s pre-trial position has emphasized that publishers had alternatives, that programmatic advertising market competition has been robust, and that the practices the DOJ characterizes as anticompetitive were standard ad technology practices responding to real operational requirements.
The Defense Strategy
Google’s defense brief, as summarized in pre-trial filings, makes several arguments worth understanding because they’ll shape the trial testimony.
On last look: Google argues that publishers knowingly used DFP, that DFP’s auction mechanics were disclosed, and that publishers who wanted to avoid Google’s auction dynamics had viable alternatives. The company also argues that the “last look” framing mischaracterizes how the auction worked technically, and that the net effect was competitive rather than exclusionary.
On header bidding throttling: Google’s defense characterizes Project Poirot as a yield optimization tool that responded to quality signals about non-Google inventory, not a deliberate competitive suppression mechanism. The company will likely present technical testimony arguing that the bid adjustments were algorithmically reasonable responses to measured performance differences.
On tying: Google argues that the AdX/DFP integration delivers technical efficiency benefits that justify any competitive effects, and that the relevant market definition the DOJ uses is artificially narrow.
Why Divestiture Would Fundamentally Reshape Programmatic
The remedy question is ultimately more consequential than the liability question for the market. The DOJ has signaled it may seek divestiture — specifically, requiring Google to separate AdX or DFP from its other advertising businesses.
If AdX were divested — sold as an independent exchange — the dynamics of the open exchange market would shift meaningfully. AdX commands approximately 40% of open exchange volume by revenue, largely because of its bundled access to Google’s advertiser demand through DV360 and Google Ads. An independent AdX without the Google demand supply would lose its structural advantage over Xandr, Magnite, PubMatic, and other SSPs that compete without Google’s demand integration.
If DFP/Google Ad Manager were divested or made mandatory to interoperate with competing exchanges on equal terms, publishers would have genuine platform choice for the first time in more than a decade. The practical result would be increased SSP competition and likely CPM uplift for publishers as they could direct inventory to exchanges offering the best clearing prices without concern about losing access to DFP functionality.
Neither scenario is a clean outcome — transition costs would be significant, and an independent AdX or DFP would take years to stabilize. But the structural effect on programmatic competition would be substantial.
What to Watch Starting September 9
The first weeks of testimony will focus on establishing market definitions — whether the relevant market is publisher ad servers broadly, or specifically enterprise-grade ad servers for large web publishers. Google’s defense requires a broad market definition; the DOJ needs a narrow one to establish monopoly power.
Testimony from independent ad tech executives and publishers will establish competitive effects from the ground. Expect testimony from Magnite, PubMatic, and major publishers who will describe their practical experience with Google’s auction practices and DFP lock-in.
The trial is expected to run through October. Closing arguments and the judge’s eventual ruling will come after that. Whatever Judge Brinkema concludes, the trial record being built over the next two months will be the definitive documentary history of how programmatic advertising’s infrastructure actually worked.
FAQ
Q: What is the difference between the DOJ adtech trial and the separate Google Search antitrust case? These are two distinct cases with different claims, courts, and remedies. The Google Search antitrust case, presided over by Judge Mehta in the District of Columbia, concerned Google’s exclusive deals with Apple and others to be the default search engine. Judge Mehta ruled Google a search monopolist in August 2024. The DOJ adtech case, in the Eastern District of Virginia, concerns Google’s advertising technology stack — AdX, DFP, and Google Ads.
Q: Will the trial outcome affect Google’s advertising products immediately? No. Even a liability finding doesn’t produce immediate changes — remedies require separate proceedings, appeals, and compliance timelines. The trial outcome will likely shape negotiations, settlement discussions, and eventually court-ordered structural changes, but the timeline from verdict to meaningful market change is measured in years.
Q: How does this trial relate to the EU’s Digital Markets Act enforcement against Google? They’re parallel tracks. The EU’s DMA enforcement targets Google’s gatekeeper behavior across multiple markets, including advertising, search, and platform access. DMA proceedings and the US DOJ trial operate independently but share some underlying factual territory — internal documents from the DOJ discovery may be referenced in EU proceedings.
Q: Should publishers or advertisers change their ad tech stack decisions now based on the trial? Smart operators should be conducting contingency planning regardless of the verdict — specifically, evaluating what their ad serving and monetization options look like if DFP’s competitive dynamics change. Publishers should know how their revenue would be affected by alternative ad server deployments. Advertisers should understand what percentage of their programmatic spend flows through Google-controlled infrastructure.