The EU Just Cracked Open Google's Search Data Vault

On 16 July 2026 the European Commission issued two sets of binding specification measures to Google under the Digital Markets Act. In plain terms: Google now has to share the Search data it collects at scale with rival search engines and AI chatbots, and it has to open Android so that competing AI assistants work as smoothly as Gemini does.

This is the mechanism that could break Google's data moat. If you run a growth, SEO or GEO team, it changes the assumption that discovery traffic flows through one dominant front door.

What actually happened

The Commission handed Google two decisions. The first covers AI interoperability on Android. The second covers Search data sharing.

The headline from the Search decision is blunt: AI chatbots offering search functionality are now eligible to receive shared data, and, subject to anonymisation, Google should share the same data it uses to optimise its own Search. You can read the detail in the Commission's guidance on Google AI interoperability and Search data sharing.

Why this matters for discovery traffic

Google's ranking advantage has never been purely about algorithms. It is about behavioural data at a scale no rival can match: what people click, what they skip, how they refine a query, what satisfies them.

That feedback loop is the moat. Sharing it, even anonymised, hands competitors and AI search products a shortcut they have never had. In my opinion this is the most consequential structural intervention in search since the Android choice screens.

The Android side: assistants get equal footing

Around 60% of EU users are on Android, and until now third-party AI assistants had restricted access to core operating system functions. The new measures let users activate their preferred assistant by voice, delegate real tasks like booking a taxi, and get in-app suggestions, all the things "Hey Google" could do exclusively.

From my observation, this is where consumer behaviour actually shifts. When a rival assistant can complete tasks natively, the default habit loosens.

Google shares the same categories of Search data it uses internally, subject to anonymisation.
A multi-layered anonymisation method applies, developed with privacy experts and aligned to the draft Joint Guidelines on the DMA and GDPR interplay.
Google can screen a recipient first for serious cyber security or data protection risk.
A fair pricing formula and a transparent access process govern who gets the data and at what cost.

The Commission has also left the door open to amend the anonymisation measures based on independent evaluation. This is a living decision, not a one-off ruling.

What it means for GEO teams

I think the practical takeaway is a diversification one. If AI chatbots can train and optimise on Google-grade signals, the quality gap between Gemini, ChatGPT, Perplexity and privacy-focused engines narrows fast.

That means your discovery traffic starts fragmenting across more surfaces, and structured, machine-readable content becomes non-negotiable. This connects directly to how the DMA forces Google to share anonymised search data under Article 6(11), and to the shift covered in how ChatGPT adoption is reshaping where audiences actually search.

The action to take now

Audit which non-Google surfaces already send you traffic, and treat AI assistants as a distinct discovery channel with its own measurement. This could be the moment the single-engine strategy stops working, so build for a market where several engines carry Google-grade data.