# EU Orders Google to Share Search Data With AI Rivals and Open Android Assistant Slots

Anonymized query, click and ranking data must reach competitors on fair terms from January 2027, with eleven Gemini-only Android features opened to rival assistants from mid-2027.

- Published: 2026-07-18T05:46:36.130Z
- Canonical: https://polylog.news/ai/2026-07-18/eu-orders-google-to-share-search-data-with-ai-rivals-and-ope
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [Polylog editors](https://polylog.news), [France 24](https://www.france24.com/en/europe/20260716-eu-orders-google-to-share-search-data-and-open-android-system-to-ai-rivals), [European Commission (DMA)](https://digital-markets-act.ec.europa.eu/commission-provides-guidance-google-ai-interoperability-android-and-sharing-google-search-data-under-2026-07-16_en)

The European Commission issued guidance on July 16 requiring Google to give rival search engines and AI services access to certain anonymized Google Search data, including [query, click and ranking signals on fair, reasonable and non-discriminatory terms](https://digital-markets-act.ec.europa.eu/commission-provides-guidance-google-ai-interoperability-android-and-sharing-google-search-data-under-2026-07-16_en). The data-sharing obligation is set to begin in January 2027.

Separately, the Commission told Google to [open eleven Android features currently reserved for its Gemini assistant](https://www.france24.com/en/europe/20260716-eu-orders-google-to-share-search-data-and-open-android-system-to-ai-rivals), so users can launch a competing assistant by voice and delegate tasks such as booking a ride. That interoperability requirement applies from the next Android release scheduled for mid-2027. Both measures fall under the Digital Markets Act, which allows fines of up to ten percent of worldwide annual turnover for non-compliance.

Google objected. Chief Legal Officer Kent Walker warned that the measures ["risk undermining privacy and security safeguards"](https://t.me/aipost/7552), the standard argument incumbents raise against data-portability mandates. The Commission's reply is that the search index and interaction logs are the durable input rivals cannot replicate, and that access, not model quality, is now the binding constraint on competition in AI-mediated search.

What is verified is the obligation and the timeline. What is not yet settled is the mechanism, meaning which data fields, at what granularity, and through what interface competitors will actually receive. Firms most likely to benefit, including OpenAI and Perplexity, have said little about how usable anonymized aggregate signals will be for training or ranking.

## What this means

The core asset in search-plus-AI is proprietary interaction data, and the Commission is trying to separate it from the model layer. If the shared data is granular enough to improve a rival's ranking or grounding, it erodes Alphabet's structural advantage in Europe through the distribution and data channel, not through model capability, and it lowers the data-acquisition cost for AI search entrants. If the anonymization strips the data of practical value, the order becomes symbolic and the advantage holds.

## What to watch

- The technical specification of what data fields and granularity Google must expose, which decides whether rivals get a usable signal or an unusable aggregate.
- Whether US regulators or courts pursue parallel search-data remedies, which would turn a European-only remedy into a global reallocation of Google's data advantage.