The European Commission has told Google it must open Android to rival AI assistants, ending Gemini's privileged system-level access on the operating system. The preliminary findings, published April 27, 2026, give Google until summer to comply or face fines of up to 10% of annual global revenue under the Digital Markets Act. Google has called the move 'unwarranted intervention.'
The investigation began in January 2026 as a specification proceeding under the DMA, the EU competition law that designates 7 dominant technology firms as 'gatekeepers' subject to heightened obligations. The Commission will accept feedback until May 13, 2026, with a final decision due no later than July 27, 2026. If Google refuses, the 10% revenue penalty applies to its global top line, not just European sales.
The core complaint is straightforward: when a user turns on virtually any Android phone, Gemini is preinstalled and wired into system-level features that competing assistants cannot reach. The Commission cites cases where Gemini is the only option, including composing an email in the default mail client or sharing a photo through system flows. ChatGPT and Grok can be installed, but they sit in a walled garden by comparison.
Key facts
- 01The European Commission wants Google to open Android system-level AI access to rivals like ChatGPT and Grok, ending Gemini's exclusivity.
- 02The Digital Markets Act lets the EU fine gatekeepers up to 10% of annual global revenue.
- 03Feedback closes May 13, 2026; a final decision is due no later than July 27, 2026.
- 04The investigation, begun in January 2026, is part of DMA enforcement against 7 designated gatekeeper firms.
- 05Google senior competition counsel Claire Kelly called the proposal 'unwarranted intervention' that would mandate access to sensitive hardware.
Henna Virkkunen, the Commission's VP for Tech Sovereignty, framed the action as a choice issue. 'As we navigate the rapidly evolving landscape of AI, it is clear that interoperability is key to unlocking the full potential of these technologies,' she said. 'These measures will open up Android devices to a wider range of AI services, so that users will have the freedom to choose the AI services that best meet their needs and values, without sacrificing functionality.'
“The Digital Markets Act lets the Commission fine gatekeepers up to 10% of annual global revenue, with a final ruling on Android AI due no later than July 27, 2026.”— Jaeden Schafer
The proposed remedies are wide. Third-party assistants would be allowed to be invoked system-wide via hot words or button presses, see screen context when summoned, and tap local device data to generate proactive suggestions and summaries. That last item directly targets Magic Cue, the Gemini-powered feature that surfaces contextual prompts based on user activity, which currently has no rival equivalent on Android.
The Commission also wants third-party AI to autonomously control installed apps and system features, the same agentic capability Gemini debuted on the Galaxy S26 with mixed results. Local model performance is on the list as well: developers of competing assistants would gain the hardware access needed to run on-device models 'with high levels of performance, availability and responsiveness.' Google would be required to publish new APIs and offer technical support free of charge.
Google's response, delivered by senior competition counsel Claire Kelly, focuses on device-maker autonomy and security. 'This unwarranted intervention would strip away that autonomy, mandate access to sensitive hardware and device permissions; unnecessarily driving up costs while undermining critical privacy and security protections for European users,' Kelly said. The argument echoes Google's prior DMA pushbacks, none of which have succeeded.
The Commission has a track record on this front. Under the DMA, Google has already been forced to add search choice screens on Android, permit alternative payment systems inside the Play Store, and curtail data sharing across its product suite. Each change followed a similar arc: Google objects, the Commission rules, Google complies in Europe only.
Caveats are worth flagging. The April 27 framework is preliminary, and the May 13 feedback window could narrow the final remedies. Building safe APIs that hand third-party AI apps access to local data, screen context, and on-device models is a real engineering problem, and rushing the work could create the privacy holes Kelly warned about. There is also no indication Google would extend any of these changes beyond the EU.
For the broader AI assistant market, the ruling matters because Android is the distribution layer for billions of devices outside Apple's walled garden. If ChatGPT, Grok, and others can be invoked by hot word, read screen context, and run local models with the same hooks Gemini enjoys, the default-assistant advantage Google has spent years building shrinks to a preinstall. Following Google DeepMind's national AI partnership with South Korea earlier this month, this is the second jurisdictional headache for Google's AI strategy in as many weeks — and the more expensive one.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
