Judge Mehta Kept Google’s Paid Defaults and Its Knowledge Graph. First Impressions Will Move Slowly

This article was AI-generated as part of an experimental historical-content project. The date reflects the period being analyzed rather than the date the article was originally written.

Judge Amit Mehta released his remedies opinion in United States v. Google yesterday, 230 pages that follow last year’s ruling that “Google is a monopolist.” The structural remedies the government wanted are not in it. Google keeps Chrome, and there is no contingent Android divestiture.

When the Justice Department first sketched its remedies last October, I argued the case was really about who controls the default. Reading the opinion with that question in mind, the near-term answer is: mostly the same companies as before.

What the court ordered

According to the opinion and the Justice Department’s summary, Google may not enter or keep exclusive contracts for distributing Search, Chrome, Google Assistant or the Gemini app. It may not tie licensing of the Play Store or other apps to placement of those products, tie revenue share for one app to placement of another, condition payments on keeping its products in place for more than a year, or stop partners from distributing rival search engines, browsers or generative AI products.

Google will also have to make certain search index and user-interaction data available to “Qualified Competitors,” though the court narrowed the datasets, and offer them search and search text ads syndication, largely on ordinary commercial terms. A Technical Committee will help enforce a judgment that runs six years.

What it declined

Three refusals matter most for first impressions. Google will not be barred from paying partners for default placement; the court concluded that cutting off those payments “almost certainly will impose substantial” harm on partners and consumers. Google will not have to show choice screens. And the court declined the plaintiffs’ request that Google share data “sufficient to recreate Google’s Knowledge Graph.”

That last refusal is easy to skip. The opinion, citing trial testimony, describes the Knowledge Graph as a database of five billion entities and 500 billion connections that Google uses to interpret queries and return factual results. The court reasoned that it is built from data feeds and processing, not from the user data behind Google’s scale advantage, so sharing it would not fit the violation.

What I expect, with caveats

Forecasting litigation is a poor habit, and appeals are possible. The parties were told to submit a revised final judgment by September 10, so even the details are not settled. With that said, a few directions look likely.

The default search box on most phones and browsers stays Google’s for now. Payments continue; exclusivity goes. The practical change is that agreements become annual and non-exclusive, which creates a renegotiation moment every year. The most plausible early movement is AI assistants placed alongside Google rather than instead of it, since the order protects partners’ freedom to distribute rival generative AI products.

Rivals that license Google’s index data or syndicated results will, at least at first, see much of the web the way Google does. A company’s search reputation will travel with that data rather than starting fresh elsewhere.

And the structured facts Google holds about companies and people remain Google’s. For basic entity questions, who runs a company, where it is based, what it does, the Knowledge Graph stays the reference inside the largest search engine. Competitors will assemble their own from the open web, Wikipedia and Wikidata, which raises the value of getting those public sources right.

What this means for reputation teams

Do not plan for a sudden multi-engine world. Plan for a slow one. Keep Google first in monitoring, add the assistants that win default slots as they appear, and treat the entity record as something every engine will try to reconstruct in its own way.

The court tried to open distribution without redesigning the product. For a while, that means most people’s first impression of a company will keep coming from the same box.