July 17, 2026, (Inside AI) — Alphabet's Google must grant OpenAI and other AI rivals access to key Android features and share anonymized search data, the European Commission ordered on Thursday. The binding specifications under the Digital Markets Act (DMA) aim to dismantle what regulators see as entrenched advantages for Google Search and its Gemini AI assistant.
The decision, issued six months after the Commission opened specification proceedings, details 11 Android functionalities that must be opened to qualified competitors. These include the ability for users to activate a rival AI assistant via voice commands—much like "Hey Google"—to perform tasks such as booking a taxi or searching for information. The changes will roll out with the next Android iteration starting July 2027.
In a separate but linked requirement, Google must share data it collects to optimize its own search services with third-party AI chatbots that include search capabilities, such as OpenAI's ChatGPT. The data must be anonymized, and the provision includes a formula to calculate the price. This obligation takes effect in January 2027.
EU tech chief Henna Virkkunen framed the move as a victory for consumer choice.
"Thanks to these measures we hope to see emerging alternatives to Google Search and Google's AI services, such as Gemini, and that users in the EU can enjoy greater choice of services," she said in a statement.
Google immediately pushed back, warning of privacy and security risks. Kent Walker, the company's president of global affairs, said in an email:
"Today's decisions risk undermining vital privacy and security guardrails for millions of Europeans. We have repeatedly offered solutions to safeguard users while satisfying the DMA's goals, but these rulings discount extensive evidence of user harm."
The Commission countered that the measures include robust safeguards. Google will only offer the 11 features to rivals that meet strict security and privacy criteria, and it can first assess whether a competitor poses cybersecurity or data protection risks before granting access.
The specifications mark the first time the DMA has been used to force an incumbent to share proprietary data and operating system hooks with AI competitors. The move reflects a broader regulatory anxiety that foundation models are becoming the new gateways to digital information, and that without intervention, the same companies that dominated web search could lock up the AI assistant market.
The data-sharing formula is likely to become a flashpoint. While the Commission says it ensures fair compensation, Google and other gatekeepers have historically argued that forced data sharing undermines investment incentives and exposes trade secrets. The requirement to anonymize data may also limit its usefulness for training competitive models, a tension the Commission acknowledges but has not fully resolved.
The timeline gives Android developers more than a year to integrate the new access points, but the real test will be whether rivals can build compelling alternatives that users actually adopt. Past DMA interventions, such as browser choice screens, have shown mixed results in shifting consumer behavior.
For OpenAI, which has been aggressively expanding its search capabilities, the data access could accelerate its ability to compete with Google on factual queries. Other AI chatbot developers with search functions, including Microsoft's Copilot and emerging European players, also stand to benefit.
The decision does not require Google to share the underlying algorithms of its search ranking or Gemini models, nor does it mandate interoperability at the model level. It focuses narrowly on data and device-level access, leaving the core competitive dynamics of AI model quality largely untouched.
Google's compliance will be closely watched. The Commission has already fined Apple and Meta for DMA violations, and it has signaled that it will not hesitate to impose penalties of up to 10% of global annual turnover for non-compliance.
