A remedy took effect
with no rules attached.

On 2 September 2025, the US District Court for the District of Columbia issued its remedies decision in United States v. Google.¹

What it rejected mattered less than what it imposed. Divestiture of Chrome was refused. The court declined most of what the Department of Justice asked for and chose behavioural remedies over structural ones.¹

Three things it did impose:

A court ordered the most valuable index in commercial history to be shared with rivals. Almost nobody in search marketing is planning around it.

The regime took effect. The terms did not.

Here is where it becomes genuinely unusual.

As of 3 March 2026, the structured data-sharing regime was triggered, with a newly formed Technical Committee working to implement it.¹

That five-member committee has not finalised the licence terms or the privacy safeguards governing how Google must share search data with rivals, leaving key operational details unresolved months after the order took effect.¹

An obligation without operative terms is not quite an obligation. A qualified competitor wanting index access today has a court order entitling them to something and no document specifying what, on what terms, with what privacy constraints, at what price.

This is not incompetence. Defining how to share a search index without transferring the personal data of billions of users is a genuinely hard technical and legal problem, and five people are being asked to solve it against no precedent. But the practical position is that the remedy is simultaneously in force and unusable.

And Google is appealing to remove it

On 20 January 2026, Google appealed, arguing the remedies are contrary to law and that the data-sharing and syndication requirements should not be in effect while the appeal is pending.²

The D.C. Circuit's docket does not yet reflect a briefing schedule, which means the appeals process is likely to extend well into 2027.¹

Why this matters to search marketing specifically

The instinct is to file this under legal news. That is a mistake, and the reason is the syndication remedy.

If a qualified competitor can licence Google's search results and text ad feeds, then the barrier to launching a credible search product collapses from "build an index over twenty years" to "sign an agreement". The thing that made search a single-destination market was never the interface. It was the index.

Three consequences follow, and none is priced into how search marketing is currently planned.

Why behavioural remedies usually disappoint

The court chose behaviour over structure. That choice has a track record worth knowing, because it predicts the most likely outcome here better than any reading of the specific terms.

Structural remedies - divestiture, separation - change what a company is. They are self-executing: once an asset is sold, no ongoing supervision is required. Behavioural remedies change what a company must do, and they require somebody to define the doing, monitor it, and adjudicate disputes about it, indefinitely.

That is precisely where this one sits. A five-member Technical Committee is now responsible for specifying licence terms and privacy safeguards that a well-resourced and highly motivated counterparty will scrutinise line by line. The committee has been at it since March and has not finished.

A remedy that requires continuous definition by a small committee, against a party appealing its existence, tends to be defined slowly and narrowly.

We are not predicting failure. We are noting that the mechanism chosen has a known failure mode - dilution through implementation - and that the observable facts so far are consistent with it. The alternative reading is equally available: hard problems take time and the committee is being careful with billions of people's data, which is exactly what it should be doing.

The definition that decides everything

One undisclosed term does more work than the rest of the order combined: who counts as a qualified competitor.

Set the bar high - established search operators with existing scale - and the remedy reaches perhaps two or three companies who least need help. Set it low and Google is obliged to share its index with any well-funded entrant, which is a materially different market.

We could not obtain the test. It is the single detail we would most want, and its absence from general coverage suggests the search marketing world has not yet registered that the question exists.

What to do about it

Track it as a market-structure input, not as news. Two dates matter: whether a stay is granted on the data-sharing obligation, and when the Technical Committee publishes licence terms. Both are checkable and neither will be reported by search marketing publications.

Do not restructure anything yet. The remedy is in force and unusable, under appeal, with no briefing schedule. Any plan that assumes a multi-destination search market in 2026 is planning against an outcome that has not happened.

Do prepare for the cheap version of it. If your organic work is heavily tuned to one destination's specific behaviours, that is a concentration risk regardless of how this case resolves. Portable fundamentals - clean rendering, accurate structured data, genuinely useful pages - survive every outcome here.

Watch who becomes a "qualified competitor". The term is defined in the order and its test determines whether this remedy produces three new entrants or none. It is the single most consequential undisclosed detail.