The raw material of search opens to rivals
Under binding decisions adopted by the European Commission on July 16, 2026, Google must share anonymised search data with rival search engines and open Android to competing AI assistants. The data covers ranking, query, click and view categories, and AI chatbots offering search functionality are among the eligible recipients. This issue covers the July 21 agenda and was published on July 25, 2026 to complete the archive.
Two binding decisions on two fronts
The Commission issued two binding sets of specifications for Google under the Digital Markets Act. The first concerns interoperability on Android: competing AI assistants should be callable by voice command, and users should be able to delegate tasks such as booking a taxi or asking about a location to alternative assistants, with privacy and security safeguards protecting device integrity. The second concerns search data sharing: the data Google collects to improve its own search service must be shared, in anonymised form, with other search engines. Eligible recipients include AI chatbots that offer search functionality.
Which data, on what terms, and when
The sharing covers four categories: ranking, query, click and view data. Free and paid results are included, along with metadata such as language and device type. Access comes on fair, reasonable and non-discriminatory terms with cost-based pricing. The Commission describes a multi-layered anonymisation method developed with privacy experts, and notes Google may assess cybersecurity and data protection risks before sharing with particular parties.
The timeline runs in stages: an application form and information page by the end of August 2026, template licences in September, datasets in November and final pricing in January 2027, with data sharing beginning in January 2027. Eligibility is defined too: companies operating a genuine search engine in the EU for two consecutive years, or newer firms above a threshold of 50 million euro in capital and 50,000 monthly EU users. The Commission assessed that Google's first compliance proposal stripped out between 90 and 100 percent of unique queries and produced no meaningful uptake. Its reasoning notes that Google Search has held above 90 percent of the EU market for decades.
The secret of search quality was the data
A search engine returns good results not only through its algorithm but by seeing which result gets clicked after which query. Rivals without access to that loop struggle to close the gap however well they engineer. That the decision also covers AI chatbots shows the argument is not confined to classic search engines: systems that generate answers want the same raw material. Sharing that starts in earnest in January 2027 could, over the medium term, make it easier for more than one strong player to persist in search and AI answers.
What it means for businesses in Türkiye
The decision binds the EU market and creates no direct obligation in Türkiye. Its effect can still seep through the products, though: when rival engines and AI assistants start working with better data, the quality of their answers rises everywhere. The takeaway for a business is that building a visibility strategy around a single search engine is increasingly risky. What can be done today is publishing content and business information so that any system can read it correctly: clear headings, text that answers questions directly, structured data, current business details and a site structure machines can reach. These serve Google and answer-generating assistants alike. In short, prepare to work independently of the door, not for one particular door.
The UNALSOFT take
Our reading is this: as competition in search grows, the winner is not whoever memorises one platform's rules but whoever structures content so everyone can read it correctly. We build sites on that principle in our web design service: clean information architecture, a structured data layer and answer-first content. Whichever engine comes out ahead, a well-written page keeps its place.
Write for every engine, not one.
Let's look at how search engines and AI assistants actually read your content.