Google’s ad tech remedies decoded: what changes, what doesn’t and what will take years 

By Seb Joseph and Jessica Davies  •  September 17, 2026  •

Ivy Liu

At the heart of it was a concern that Google controlled both sides of the pipe ad money flows through, the tool publishers use to sell their ad space and the marketplace where those ads are actually bought, and used that control to rig the game in its own favor. Advertisers and publishers couldn’t avoid Google even if they wanted to, and Google steered the money toward itself at their expense. 

Judge Brinkema agreed with the government, and arguably with the industry, on those concerns. She didn’t agree with them on how to unwind it. They wanted Google broken up — its ad exchange sold off and the auction logic powering its ad server opened to the public. They got, in the judge’s own words, “behavioural remedies” instead, and they apply everywhere Google’s ad tech operates, not just in the U.S. Those remedies fall into three groups: rules forcing Google to open up and share data with rivals, rules stopping Google from favoring its own tools over everyone else’s, and rules stopping Google’s ad buying arm from gaming the auctions for its own ends. 

None of that will surprise anyone who’s been following the case. It was pretty much clear from the trial itself where the judge would land. What people were watching for was what the remedies actually said, and how fast they could be enforced. On paper, they fix some of the long-standing issues on the sell-side. They don’t, however, touch the rest of it. Google’s monopoly was never just about auction mechanics. It’s about the company’s ability to steer and control demand through inventory it owns and data nobody else can see. 

Here’s how the industry got there and why. 

The two behavioral remedies the ad industry actually wanted are in there: Google has to let Prebid, the open-source tool a lot of publishers already run their auctions through, pull real-time bids out of its own ad exchange (AdX). DfP gets the same treatment. In plain terms, it means publishers can finally make AdX compete for their ad space through the same neutral pipe they use for every other seller, instead of Google’s own exchange getting access nobody else can see or check. 

This was the one ad tech execs, speaking before the remedies came our, said would actually matter. AdX would start losing auctions to competitors, they said, and publishers would earn more because Google could no longer charge whatever it wanted without a real fight for business.

Not everyone buys the optimism: Others are less convinced that forcing AdX to compete through neutral pipes like Prebid would finally shift more money to publishers. Privacy expert Alan Chapell, founder of Chapell and Associates, argues that while these remedies look good on paper, Google is “brilliant at playing the behavioral remedies game.” It can surround access to AdX demand with conditions and audits, write ultra-narrow promises that ban “yesterday’s” self-preferencing while it invents new tactics, and lean on opaque privacy tech so publishers can’t really audit what’s happening to their data. The risk, in his view, is that publishers end up with the appearance of tougher rules but only “breadcrumbs” of real competitive change. 

“Once the judge decided not to push for a Chrome divestment, I was pretty confident that we were cooked,” he said. 

Dv360 didn’t make it: This is a big part of why people are skeptical of these remedies. Google argued from the start that the buy-side of its ad tech business — AdWords and Dv360 — should be left alone, since the original ruling never found Google had a monopoly in buying tools. Judge Brinkema partly agreed and partly didn’t. She zeroed in on AdWords and left Dv360 alone. Her reasoning being that AdWords specifically was part of how Google rigged the game. It steered AdWords advertisers toward bidding into AdX specifically, and made publishers run AdX and DfP together if they wanted a shot at that AdWords money. Dv360 got left out because nobody proved in court that it worked the same way. The government never put forward the evidence to show Dv360 played a part in the original conduct so they just had nothing to base a remedy on.  

Bottom line: Regulating auction access while leaving the buy-side integration intact doesn’t open the market; it just turns Google into a supervised public utility and sets up years of algorithmic hide-and-seek.

It’s going to take a long time to show up: Google has 12 months to open up AdX to rival ad servers and 12 to 15 months to connect AdX and DfP to prebid. Judge Brinkema has said the whole judgement should be up and running within 15 months. That’s over a year before any of this shows up in what publishers actually earn, or which company wins more auctions. 

One ad tech specialist, who has examined Google’s proposals and asked to remain anonymous to speak freely, said the current remedies aren’t built to improve publishers’ revenues at all. The legal commitments aren’t designed to put more money back in publishers’ pockets; they’re written to satisfy narrow legal promises, not guarantee any change in outcomes. In practice, he expects very few publishers to change anything, and therefore, very little to change in their bottom lines. “The commitments…they do not say that there is going to be a market impact as a deliverable,” he said. 

How this actually gets enforced is what worries people most: Before the remedies were ratified, execs said none of it would matter if breaking a rule took too long to punish. The worry was that if Google bends a rule, it could take eight to nine months to drag a complaint through the courts, by which time the damage is done. Yes, there will be an independent monitor watching Google that it has to pay for, but the judge sided with Google on how that monitor works. It reports every three months instead of whenever something goes wrong, and Google gets over a month to respond and fix things before anyone can even take it to the judge. That’s the slower, more Google-friendly version of what was on the table. 

The remedies only cover the part of the market that’s shrinking: All of this only applies to open web display ads — the traditional banner and display ads people see on sites. It does nothing for the fast-growing areas like streaming TV ads, in-app ads or retail media ones on sites. Google’s own numbers make this point: the share of AdWords ad impressions going to that traditional display category fell from over 40% to just 11% between 2019 and 2025. Meanwhile, those other formats have been growing. So the fix targets the part of the pie that matters less every year. 

A workaround nobody was watching for: The court pointed out that Google already has a tool called Google Partner Bidding, or gBid Direct that lets its ad buying arm bid directly into a publisher’s auction for in-app mobile ad inventory, skipping the exchange altogether. It doesn’t do it for open web display yet but the court said it could and treats that as a real risk worth watching.