Doubao AI Phone Sparks Privacy Debate at MWC with System-Level Access
The AI Phone That Sees (and Does) Everything
At this year's Mobile World Congress in Barcelona, while foldable screens and satellite tech drew crowds, it was an unassuming preview device called the "M153 Doubao AI Phone" that became the talk of the show. This ByteDance-ZTE collaboration demonstrated something unprecedented: an AI assistant that doesn't just suggest actions, but actually performs them across your apps - tapping buttons, retrieving data, and navigating interfaces like a ghostly digital butler.
The God Mode Controversy
"It's like giving your phone X-ray vision," remarked one attendee after watching the demo. The Doubao phone achieves this by requesting what security experts call "God mode" permissions within Android - access levels typically reserved for system developers. This allows it to read screens, simulate touches, and extract data across app boundaries.
Tencent CEO Ma Huateng voiced stark concerns during a panel discussion: "When an AI can not only see your messages but reply to them, when it can not only view your calendar but schedule meetings - where do we draw the line between convenience and surveillance?"
Permission Paradox
The Doubao team emphasizes their strict opt-in authorization process. "Every action requires explicit user consent," claims their lead developer. But compliance specialists point out a troubling reality: most users blindly accept permissions without understanding the implications.
Privacy advocates highlight three core issues:
- Data spillover: When AI accesses multiple apps, it creates unseen data bridges between previously isolated services
- Platform conflicts: App developers may resist such deep integration that bypasses their official APIs
- The accountability gap: If an AI mistakenly transfers money or shares private photos, who's liable - user or developer?
The New Frontier of Mobile AI
The Doubao experiment reveals how quickly we're approaching uncharted territory in smartphone evolution. As Beijing Chuntian Zhiyun Technology (ByteDance's AI arm) pushes these boundaries, they're forcing the industry to confront fundamental questions:
- Should phones have a standardized "AI permission" framework like camera/microphone access?
- How do we prevent AI capabilities from becoming surveillance tools?
- Can ecosystems maintain security while allowing this level of cross-app automation?
This isn't just about one phone - it's about defining the rules for the next generation of mobile devices. As one analyst put it: "We're not just building smarter assistants anymore. We're creating digital entities with unprecedented access to our digital lives."
Key Points:
- The Doubao AI Phone demonstrates system-wide automation capabilities at MWC 2026
- Requires high-level Android permissions that concern privacy advocates
- Tencent CEO publicly questions the security implications
- Raises new questions about liability and platform boundaries in the AI era



