Google trial: Chrome is safe, but data will be open to rivals
Chrome is safe and there will be no breakups, even though Google will no longer be able to sign exclusive agreements and will have to share some of its search data with qualified competitors. This is the final decision of Federal Judge Amit P. Mehta to correct the monopoly in online search, which has already left competitors and critics dissatisfied.
Chrome and Android remain in the hands of Alphabet, therefore, and billion-dollar payments to Apple continue in order to keep Google as the default search engine on iPhones. The only substantial change is that Google will have to open up a package of index data and interaction signals used to build SERPs, with limited sharing and privacy protections.
Mountain View disputes the impact on privacy and quality, while several observers speak of a “light” remedy. For those working on Search and AI, however, the perimeter is changing: exclusives are ending, and datasets and syndication are coming in, which can accelerate credible alternatives.
Judge Amit Mehta’s decision
Tuesday, September 2, 2025, was a “historic” date for the digital world: after several months, the ruling on the “antitrust case against Google” was handed down by the Washington District Court, signed by Federal Judge Amit P. Mehta, the same judge who in August 2024 had already found Google guilty of monopolizing the online search market. After another year of debate on corrective measures, yesterday the remedies were set: no drastic breakups, but limits on exclusive contracts and obligations to share data with competitors.
Mehta explained that the goal is not to retroactively punish Google, but to ensure that its dominance does not extend to new frontiers of search, particularly those based on generative artificial intelligence. His choice, therefore, fell on a compromise: preserving the economic balance of partners such as Apple and Mozilla, which receive billions a year to set Google as their default search engine, while at the same time opening up the market to greater competition.
Concrete remedies
Essentially, the judge ruled that Google can maintain control of Chrome and Android, but can no longer lock in the distribution of Search and AI services through exclusive agreements with partners such as Apple, Samsung, or Mozilla.
It will also have to share with qualified competitors some of the data collected in online searches, which the court considers an essential ingredient in enabling other operators to develop credible alternatives.
More specifically:
- Payments for defaults. Google will be able to continue paying partners such as Apple or Mozilla to remain the default search engine, but with specific limits. Contracts may not exceed one year in duration and may not contain clauses that prevent the partner from also hosting competing engines. This means, to put it bluntly, that Google can still buy the “front row seat” on Safari, but without clauses that exclude rivals and with restrictions on the duration of defaults.
- Exclusive contracts. Agreements that guaranteed Google exclusive distribution of Search, Chrome, or GenAI services (Assistant, Gemini) are prohibited. In practice, smartphone and browser manufacturers will be able to accept Google’s offer, but will be free to integrate alternatives as well.
- Data sharing. The company will have to open a selected package of datasets to “qualified competitors”: reduced snapshots of the search index and user interaction signals useful for building and refining a SERP. Advertising data will remain excluded and will not be shared.
- Syndication. Google will have to offer search results and text ad syndication services, allowing competitors to rely on its system while they develop their own. The economic conditions are not set by the judge, but must remain consistent with existing commercial practices.
- Governance and controls. A five-member Technical Committee has been established, with expertise in privacy and data security, to oversee compliance with the measures, set limits and standards for sharing, and authorize any audits.
Why this choice
In his reasoning, Mehta writes that forcing the sale of Chrome or Android would have been a “poor fit” (a disproportionate measure) for the case, too risky for users and difficult to manage technically. He preferred to focus on “forward-looking” remedies: blocking exclusive agreements and opening up portions of the index and interaction signals, so as to promote competition without compromising the quality of the service (and without dismantling Google’s structure).
A key passage concerns artificial intelligence, a factor to be monitored to prevent the same imbalances from transferring from traditional search engines to generative systems. “The emergence of GenAI products has changed the course of this proceeding,” the decision reads; For the judge, the risk is that the imbalance already established in traditional search will be replicated in AI-generated responses. For this reason, in addition to limiting exclusive contracts, he ordered the sharing of targeted datasets (including signals such as Glue and RankEmbed) precisely to give oxygen to new operators in the field of AI search.
Google’s position
In an official statement released immediately after the ruling, Google welcomed the fact that there are no breakups or blanket bans on payments for defaults, but strongly contested the obligation to share search data. In particular, it warned that opening up user-collected signals to third parties risks compromising privacy and the quality of results, insisting that competition is already intense thanks to platforms such as TikTok, Amazon, and ChatGPT.
It also pointed out that Chrome and Android were not built to operate as standalone businesses and that a ‘break-up’ would do more harm than good. ‘We will continue to focus on products that people choose and love,’ Mountain View wrote.
What changes for search and AI
The decision does not break up Google’s empire, but it does redraw its boundaries. The ban on exclusive clauses means that smartphone and browser manufacturers will no longer be obliged to lock in Google as their only option: they will be able to enter into parallel agreements with other search engines or artificial intelligence services.
It is not a total opening, but for Bing, DuckDuckGo, or new AI platforms, it is the first real opportunity to enter devices and distribution channels. At the same time, the obligation to share certain search data gives competitors a more solid starting point: they will be able to access snapshots of the index and interaction signals that can be used to improve crawling and ranking.
The datasets to be shared
The heart of the remedy is precisely the obligation to make certain search data accessible. This does not include Google’s entire index or knowledge bases about people, places, and things, or advertising data, but rather selected packages that can still make a difference:
- Reduced snapshot of the search index: includes unique document identifiers (DocID), map between DocID and URL, date of first sighting of a page, last crawl performed, spam scores, and device type flags. This information, when added together, helps to understand what was seen by Googlebot and when, facilitating the construction of an updated and less redundant index.
- User interaction signals: Google will have to open up parts of the systems that build the SERP based on behavior to qualified competitors. This is where Glue (the set of clicks and interactions that determine the perceived relevance of results) and RankEmbed (semantic embedding models that map similarities and context of content) come into play. In other words, rivals will receive a fragment of the “cognitive engine” that refines rankings.
Sharing will be contingent: not continuous flows, but limited access, with caps defined by the Technical Committee to balance utility and privacy protection. In practice, competitors will have more oxygen, but they will not be able to replicate the complex system that has made Google dominant.
Why it matters to competitors
For Microsoft, OpenAI, or startups like Perplexity, being able to access this data means accelerating the construction of alternative engines without starting from scratch. They will not have the same scale as Google—which collects nine times more data than its rivals every day, and this difference in volume cannot be negated by a court order—but they will at least have a credible starting point for refining crawling, indexing, and ranking. The absence of advertising data limits commercial reach, but on a technical level, the impact can be significant.
The risk of a monopoly in AI
In his reasoning, Mehta explicitly links the remedy to the rise of generative AI. The goal is not only to rebalance classic search, but to prevent Google’s advantage from transferring unchecked to Gemini and future assistants integrated into devices. Hence the decision to prohibit exclusive agreements for AI products as well and to require signal sharing. It’s a way of saying: the AI search game cannot start with a winner already designated.
Initial reactions: markets, government, and Google
The stock market immediately rewarded the decision: Alphabet’s stock rose more than 7% in a few hours, a sign that investors saw the remedies as a favorable compromise. Apple also gained more than 3%: Google’s billion-dollar payments to be the default search engine on iPhones were not questioned, ensuring Cupertino a stable share of its profits.
The Department of Justice spoke of a “significant victory,” claiming to have imposed concrete limits on Google’s monopoly. At the same time, antitrust chief Gail Slater left the door open for an appeal to seek tougher remedies, implicitly acknowledging that the ruling does not go as far as the government had requested.
For its part, Google has, as mentioned, chosen a balanced approach: on the one hand, it expressed relief at the absence of structural remedies, while on the other, it strongly contested the obligation to share search data.
Competitors’ skepticism
Direct rivals were much more critical. DuckDuckGo called the decision ineffective, because Google will continue to have an unattainable scale of data and will still be able to pay partners to remain the default option. Other emerging operators, such as Perplexity and OpenAI, pointed out that reduced snapshots and partial signals are not enough to rebalance an advantage built up over twenty years of data collection.
The academic and antitrust front
The harshest voices came from competition associations, activists, and scholars. The American Economic Liberties Project called the ruling a gift to an already convicted monopolist, with the famous comparison: “It’s like convicting a robber and asking him to write a thank-you note.” Matt Stoller accused Judge Mehta of failing to respect the very principle he cites in the ruling—that of “ending the illegal monopoly”—by leaving Google’s power structure intact.
The impression from the SEO world and the digital community is similar: the ruling does not affect Google’s centrality, but it does create margins that need to be watched carefully. The sharing of data such as Glue and RankEmbed is seen as a significant step, albeit a limited one, because it could help competitors build alternative ranking systems. But the general feeling is that the verdict is not enough to rebalance the market.
Who wins and who loses
In the short term, the winners are Google and Apple, which retain their assets and economic flows intact. The losers are their competitors, who only gain limited access to data and no immediate redistribution of market share. However, there is a new element in the background: for the first time, the court has set a ceiling on Mountain View’s bargaining power. The game is not over yet and will also be played out in terms of future implications, from the AI Overview to new search methods that could reduce Google’s centrality.
What this means for those working in digital
The ruling does not dismantle Google, but it does force it to change some rules. For those working in SEO and digital marketing, this means two things.
The first one is that exclusives are no longer possible: manufacturers and browsers will have the freedom to open up to alternative solutions, and in the coming months we may see tests and partnerships that would have been unthinkable today. The second is that, for the first time, Google’s competitors will have access to index data and behavioral signals such as DocID, crawl, spam score, Glue, and RankEmbed: it is not the entire search engine, but it is material that can feed new ranking systems and reduce the gap, at least in part.
For us professionals, this scenario does not change the immediate centrality of Google: most traffic will continue to pass through it. However, the competitive landscape is changing. If rivals make good use of the datasets they receive and the syndication granted, search could become more pluralistic, and with it, the points of access to online visibility. This is a development to monitor, because it affects both traditional SERPs and the generative systems that are emerging in AI engines.
In other words, we are not facing “game over,” but a game that is opening up on multiple tables. For those involved in SEO and content strategy, this means keeping an eye on the new distribution dynamics and how Google’s competitors will be able to use the data made available.
Because the real question, from today onwards, is no longer just how to climb Google’s SERPs, but how to prepare for a market in which search and AI can be contested by multiple players.
