The European Commission just drew a line in the sand: Google’s stranglehold on Android AI is over. After months of investigation under the Digital Markets Act, regulators have formally proposed forcing Google to open up Android to rival AI assistants. This isn’t a suggestion. It’s a regulatory ultimatum that could land by July 27, with fines up to 10 percent of Google’s global revenue if they drag their feet. The core complaint? Gemini isn’t just pre installed. It’s baked into the system at a level no third party can touch.
The Uneven Playing Field
Right now, Gemini gets exclusive backstage passes to Android’s core functions. Want an AI to read your screen, send an email from your default client, or trigger a proactive suggestion based on your calendar? That’s Gemini territory only. The Commission’s framework demands that third party AI services like ChatGPT, Grok, or smaller European models get the same system level privileges: hot word activation, screen context access, local data for summaries, and even the ability to autonomously control apps. Google argues this is an ‘unwarranted intervention’ that would force them to hand out sensitive hardware permissions. But the real story is simpler: Google built a walled garden, and the EU is bringing a wrecking ball.
APIs, Access, and the Local Model Mandate
Beyond just surface level integration, the Commission wants to force Google to create new APIs and offer free technical assistance to competitors. The most ambitious demand involves local AI models. Google has been slow to let anyone else tap into on device processing for features like Magic Cue. The EU wants a mandate guaranteeing developers ‘high levels of performance, availability and responsiveness’ for running models locally. This is a direct challenge to Google’s strategy of using hardware advantages to lock in Gemini. The public comment period closes May 13. If Google loses this fight, Android in Europe could become the most open AI platform on the planet. And it’s about time.
Source: Arstechnica
