The Filing That Opened the Last Chapter

Late Tuesday, July 28, 2026, the Department of Justice filed its 144-page brief at the D.C. Circuit. It is not a legal formality; it is the opening salvo in the final fight over whether Google's search monopoly gets to define the AI era the way it defined the mobile era. The DOJ wants the court to affirm Judge Amit Mehta's 2024 monopoly finding and the data-sharing remedies — and, most importantly, to vacate Mehta's refusal to ban the massive default-payment deals that keep Google's competitors out of distribution.

Let me be direct: this is not a courtroom drama. It is an infrastructure story about who controls the data and distribution AI search needs to compete. The remedies Mehta ordered in December 2025 were designed for 2024-era search, and the DOJ itself admits the AI window is closing. If Google keeps its default deals and its data monopoly, no amount of clever AI startups will matter. The moat is paid for with 26 billion dollars a year in default payments.

Courthouse News reported the filing on July 29. The case began in October 2020 and is the biggest tech antitrust fight since United States v. Microsoft in 2001 — a precedent that looms large. Microsoft upheld liability but rejected a breakup, settling with lighter penalties. The DOJ is not repeating that. No breakup of Chrome or Android; instead, something more surgical: a ban on the default-payment deals that lock in Google's position.

The AI Argument at the Heart of the Case

The DOJ's brief makes an argument unthinkable in 2020. Quoting directly from the filing: 'Emerging generative AI products pose a potential threat to Google's dominance, and a competitive window may close if Google can continue using its monopoly profits — fruits of its monopolizing conduct — to squelch the threat.'

Read that again. The DOJ is not just arguing about the past; Google's monopoly profits are buying time in the AI transition. The window for competition is open now, but it will close — and Google is using 26 billion dollars a year in default payments to keep it shut.

This is the 'exclusionary cycle' argument. Google's exclusive deals foreclosed half the search market for years, depriving rivals of data and scale, and the profits bought more exclusion. The DOJ says the loop now applies to AI: Google is not just defending its search monopoly, it is extending it into the next generation of search technology.

I have run servers and built businesses on platforms that could crush me. This is not abstract theory: Google's search index and clickstreams are the raw material for AI training and inference.

The Remedies Fight: Data, Defaults, and 26 Billion a Year

Current state of play: Mehta's December 5 order approved a five-member Technical Committee to oversee sharing of Google's search index and user-interaction data with competitors including OpenAI and Perplexity, plus syndication of search results and ads. The remedies took effect February 3, 2026; exclusive default agreements must end by June, replaced by chooser screens.

But here is the critical gap: Mehta rejected the DOJ's request to ban the default-payment deals outright, just as he rejected the September 2025 request to divest Chrome and Android. The DOJ declined to appeal that decision. Instead, they seek the more effective remedy: a payment ban on default agreements.

Why is this the right move? Because divestiture is a blunt instrument: years to implement, massively disruptive, and it often fails to change the underlying economics. A payment ban is surgical — you cannot pay distributors to be the default. You have to win the choice screen. You have to compete on merit.

Google pays over 26 billion dollars a year to distributors. That is not a rounding error; it is a toll booth on the entire search market. Every time you open Safari on an iPhone or Firefox on a laptop, Google paid for that privilege. That is not competition. That is rent. The DOJ's cross-appeal asks the D.C. Circuit to close this gap; Google's own appeal requested a pause on the remedies. The DOJ is saying: no, the remedies are not enough. We need the payment ban.

Google's Defense: Fair and Square, Open Source, and Trade Secrets

Google's position is predictable and, in some ways, defensible: they won the market 'fair and square,' the remedies would expose trade secrets and hand windfalls to AI developers like OpenAI that were never harmed, and a default-payment ban would threaten partners like Mozilla and raise smartphone costs.

Google's 111-page opening brief, filed May 22-23, argues that Mehta 'piled error upon error' and confused harm to competitors with harm to competition. That is the classic antitrust defense: competitors hurting does not mean competition is hurting. Legitimate argument — and, in this case, wrong on the facts.

On trade secrets: Google says sharing its search index and user data would expose proprietary information. That is true. But the Technical Committee manages that risk — five members, appointed by the court, overseeing the sharing. Not a free-for-all; a controlled process.

The Mozilla argument is more interesting. Mozilla relies on Google's default payments for a significant portion of its revenue, and a payment ban would hurt it. A real concern. But the answer is not to preserve a monopoly to protect a dependent partner. Mozilla has had years to diversify and has not done so.

And 'fair and square'? Google built a better search engine in the early 2000s — true. But the market has not been competitive for over a decade; the default deals foreclosed half of it. You cannot win a race when you built a wall across the track.

The Questions Nobody Has Answered

There are three questions nobody has answered satisfactorily.

First: What is the actual competitive harm in the AI era? The DOJ says Google will use its monopoly profits to squelch AI competition. Google says AI is a new market and they are competing fairly. Google has the data, the distribution, and the capital. Is that a competitive advantage or an illegal moat?

Second: What is the right remedy for a market changing in real time? The remedies were designed for 2024-era search. A chooser screen for traditional search is fine, but what about AI assistants? What about voice search? The remedies must be forward-looking, not backward-looking.

Third: What happens to the data? Data is not static; it is generated every second, by every user, on every device. The sharing mechanism must keep pace, or the remedy becomes obsolete the moment it is implemented.

These are infrastructure questions, and the D.C. Circuit is not equipped to answer them. That is why the Technical Committee and the payment ban matter: the court sets the boundaries, and the technical experts work within them.

What This Means: The Moat Is the Story

Here is my straight-talk take. The moat is the story. Google's search business generates over 90 billion dollars a year and holds roughly 90 percent of U.S. search. That is not a market position; it is a fortress — funded by 26 billion dollars a year in default payments.

The DOJ is right to focus on the payment ban. It is the single most effective remedy available: no divestiture, no breakup. It simply says you cannot pay to be the default. You have to win the choice.

Will that be enough? Maybe. Maybe not. OpenAI and Perplexity are building real products — right now, on top of Google's data, thanks to the Technical Committee. If the payment ban is upheld, they have a real chance on distribution. If not, they will be stuck where every Google competitor has been for two decades: technically superior, commercially irrelevant.

The D.C. Circuit's ruling will set the boundaries of antitrust remedies in platform markets for a decade. It will decide whether the AI era starts with real competition or with the same default locks, just wearing a chatbot costume.

What Comes Next

Google's reply brief is due September 29, 2026, and oral arguments will come in the D.C. Circuit's 2026 term, which starts in September. The timeline is tight. The stakes are enormous.

Here is what I am watching. First, the oral arguments: the judges' questions will reveal how they think about the AI argument. Serious engagement with the DOJ's 'competitive window' theory favors the government; dismissing it as speculative favors Google.

Second, the Technical Committee. If the committee works well, it will demonstrate that data sharing is feasible without destroying trade secrets. If it fails, Google will argue the entire remedy regime is unworkable.

Third, the market. If OpenAI and Perplexity gain real share over the next six months, the DOJ's argument gets stronger; if Google's AI products maintain dominance, it gets weaker. The market will vote before the court does.

The window is closing. The DOJ knows it. Google knows it. The question is whether the D.C. Circuit does. The answer will come in the next few months — and it will determine whether the AI era starts with competition or with the same old moat, just deeper and wider.

— Allan Ali, Sylt.ing