At the 2026 Mobile World Congress in Barcelona, the spotlight was not only on foldable screens and satellite communication, but also on a preview version of a phone called "M153
According to on-site reports, Tencent CEO Ma Huateng expressed great caution when commenting on this product, publicly raising concerns about its security. The core of the controversy lies in the fact that the Doubao AI phone needs to obtain high-risk permissions within the Android system to handle tasks across apps like humans. This means that the AI can not only see your screen, but also click buttons and retrieve data on your behalf. Is this "overstepping" operation a tool for improving efficiency or a privacy black hole lurking deep within the system?
In response, Doubao emphasized that all operations strictly followed user authorization. However, according to compliance experts, the situation is not that simple. When AI starts to "take over" the phone across apps, it involves not only the risk of privacy leaks, but also the potential to touch the interests of other app platforms, leading to conflicts between data security and platform openness.
The attempt led by
The case of Doubao phone is like a mirror, reflecting the collective anxiety of the industry in pursuing "system-level intelligence": before we completely hand over our phones to AI assistants, who will guard our digital boundaries? This stir in the underlying ecosystem has just begun.
