After Friday’s proceedings, she has now retired to her chambers to mull over a decision over the coming weeks, but for many, a complete divestiture of Google’s ad exchange or ad server (i.e., the entity formerly known as DoubleClick For Publishers within Google Ad Manager) appears increasingly unlikely.

During closing arguments, the judge repeatedly steered the discussion away from theory and toward hard practicalities, indicating that pragmatism, more than pure ideals, will (likely) drive the final outcome. This is because Brinkema’s questions consistently pressed the DOJ on implementation timelines, commercial feasibility, and the (what many courtroom attendees perceived as) a lack of concrete detail in its structural remedy plan.

“Time is of the essence,” she said at one point, citing the rapid pace of technological change in advertising and AI. Any remedy, she stressed, must be “concrete” and must account for the “commercial reality” of execution.

That framing dominated the morning. While DOJ attorney Matthew Huppert reiterated the government’s preference for a structural remedy — divesting AdX, open-sourcing auction logic, or potentially forcing a sale of GAM — he conceded DOJ was “not dogmatic about the details.” But when pressed on specifics, including whether a structural remedy could be implemented during an inevitable appeal, the DOJ offered little. Judge Brinkema, noted that behavioral remedies “could go forward whether there is an appeal pending or not,.” She added, “[Divestiture] is a dramatic change,… [it] would most likely not be as easily enforceable while an appeal was pending.”

Google’s Karen Dunn capitalized on that hesitation. Dunn argued the DOJ “did not meet the heavy burden required for a divestiture,” arguing that witnesses for the defense asserted that Google’s proposed remedies would address their concerns. She further outlined Google’s opinion that the DOJ failed to prove behavioral remedies were insufficient; that a technically and commercially workable buyer existed; and that divestiture would restore competition or avoid harms to consumers.

During rebuttal, DOJ attorney Julia Tarver Wood conceded that a buyer has not been identified –although Digiday sources have relayed how loose-lipped attendees at this year’s Advertising Week New York voiced their intentions to take a look at the DFP – while the defendant argued a sale would be “messy, risky and long, with no guarantee of success.”

What the ruling will likely look like

Taken together, Brinkema’s questions point to a strong likelihood that she will opt against ordering a divestiture — at least not immediately. Her emphasis on timelines, enforceability during appeal, technical feasibility and the need to avoid abstract reasoning suggests she is leaning toward a behavioral-first remedy, potentially supplemented with guardrails or phased requirements. A ruling is expected “likely next year,” she said — most likely in Q1 or Q2 of 2026, based on her remarks.

Settlement? Appeal? Hw long will this drag on?

Brinkema remarked that she “could see a settlement” that makes sense, though she acknowledged the widening gap between the parties and the existence of “looming settlements” in other ad-tech cases. Realistically, the chance of an agreed resolution appears low.

An appeal, however, appears inevitable. Google has signaled as much, and Brinkema repeatedly referenced the complications of crafting remedies that must survive scrutiny. A full appellate cycle could take 18–30 months, pushing any structural remedy (if ordered) well into 2028.

In other words, the case may be nearing its first final judgment, but the industry is only entering the next chapter.

What we’ve heard

“We’re entering an odd moment in AI where a huge number of “AI companies” are basically wrappers on top of ChatGPT, Claude, Gemini, or open models like Mistral and Qwen. They don’t own the tech, the data, or the infrastructure — yet they’re being valued as if they do.”

– One Digiday source casts a healthy dose of cynicism among the AI hype-cycle

Numbers to know

• $1.98 billion: the amount Adobe paid to acquire SEO outfit Semrush.
• $493.8 million: the amount Microsoft received in revenue share payments from OpenAI during 2024.
• 30%: the cuts brands expect to make in their open web display ad spend due to the rise of AI search.
• 56%: the wage premium workers listing AI skills can earn, up sharply from +25% the prior year.

What we’ve covered

• How Amazon aims to do more with less

Sources suggest Amazon’s latest round of mass layoffs cut headcount in its advertising division by up to 20%, though staff there are paid through the last week of January, making it difficult to establish exact numbers.

• Pitch Deck: How Amazon plans to turn Q3’s $17 billion ad haul into Q4’s next big DSP push

A new pitch deck making the rounds at agencies drives that point home.

What we’re reading

Google boss Sunar Pichai warns ‘no company immune’ to AI bubble bursts

The most powerful chief executive in the industry says the quiet part out loud.

The EU considers cracking down on Big Tech’s cloud power

Cloud infrastructure is increasingly part of the consideration of advertisers and their supply chain, and that’s not lost on regulators.

Advertisers see now value in news, but Google knows society will pay

Richard Gingras, Google’s longtime news policy lead, he continues to push back on the Google-owned web argument, in conversation with Ricky Sutton and Alan Chapell.

Yahoo is testing 6 AI agents for advertising, leaked slides show

Yahoo hopes to capitalize weariness over Google’s market leadership and the escalating standoff between Amazon and The Trade Desk to further its marketshare of its demand-side platform.