The EU’s Antitrust Hammer Drops Again
The European Commission has completed its specification proceeding into Google’s AI practices on Android, and the verdict is damning: Google’s Gemini gets system-level privileges that no third-party AI can touch. Sending an email via voice command? That’s Gemini only. Sharing a photo through a contextual suggestion? Gemini again. The Digital Markets Act (DMA) was designed to break exactly this kind of gatekeeper stranglehold, and the EU is now demanding Google open up its OS to rival AI assistants like ChatGPT and Grok. Google’s response — calling it “unwarranted intervention” — is the predictable bleating of a monopoly that has no interest in ceding control.
What the EU Actually Wants
The proposed changes are sweeping and specific. The Commission wants third-party AI tools to be invocable system-wide via hot words or button presses, and to access screen context and local data for proactive suggestions — exactly what Gemini’s Magic Cue does today. Even more provocative: the EU wants to force Google to let other AI services autonomously control installed apps and system features, a capability Google just debuted on the Galaxy S26 (and badly botched). To make local AI models work, Google would need to provide hardware-level access APIs, technical assistance, and do it all for free. The feedback window closes May 13, with a final decision by July 27. Fines for non-compliance can hit 10 percent of global annual revenue.
Google’s Privacy Pretext Won’t Hold
Google senior competition counsel Claire Kelly argued this would “strip away autonomy” and “undermine critical privacy and security protections.” Translation: we don’t want competition. Google has a long history of using privacy and security as a shield against antitrust enforcement — exactly as Microsoft did in the browser wars — while simultaneously vacuuming up user data itself. The reality is that the DMA has a proven track record: Google was forced to add search choice screens, allow alternative Play Store payments, and limit cross-service data sharing. None of those changes caused privacy Armageddon. The same will hold true for AI. This isn’t about protecting users; it’s about protecting Google’s AI fortress.
Source: Arstechnica
