EU Orders Google to Share Search Data, Open Android to AI Rivals

The European Commission's two binding orders against Google, reported by The Verge this week, put a specific mechanism and a specific clock behind an ambition that had until now been mostly rhetorical: making room for competitors to Google Search and Gemini inside a stack Google effectively controls.

The first order, under DMA Article 6(11), requires Google to give rival search engines and AI chatbot providers access to anonymised ranking, query, click and view data from Google Search on fair, reasonable and non-discriminatory terms, starting January 2027. The second, under Article 6(7), requires Android to give third-party AI assistants the same interoperability Gemini enjoys: voice activation, system-level screen context, access to core apps like Gmail and Calendar, and the right to be designated as default. Those changes are due to reach users from July 2027.

Why this matters for the shape of the AI assistant market: the reason ChatGPT, Claude and similar services have not felt native on Android is that the hooks Gemini uses were not exposed to anyone else. Search data sharing is the more consequential half, because a general-purpose chatbot's answer quality is bounded by the freshness and breadth of the web signal it can see, and Google's index and click data are among the most valuable grounding signals in the world. Henna Virkkunen, the EU's tech chief, framed the goal plainly, saying she hopes to see "emerging alternatives to Google Search and Google's AI services, such as Gemini," and that EU users "can enjoy greater choice of services."

Google's public response, as reported by Punch, was that the measures "introduce unprecedented risks to user privacy, device security, and national security," and that they could expose Europeans' private searches to unfamiliar companies without adequate anonymisation. Take that as the company's position, not a settled finding, because the anonymisation method itself is one of the specifics the Commission still has to nail down.

The honest caveat is that the reporting does not tell you what the FRAND pricing will look like, which rival AI providers will qualify to receive the data, or how the Commission plans to police re-identification. Non-compliance fines can reach 10% of Google's global turnover, so the enforcement teeth are real, but the interesting fight over the coming months is technical, not rhetorical. If the specifics land in a workable place, the companies best positioned to benefit are the consumer-assistant builders who lack their own web index, like OpenAI, Anthropic and Perplexity, and any European search startup that until now could not afford to crawl at Google's scale.