Google’s Gemini AI is edging closer to controlling your Android phone screen, with new beta code hinting at “screen automation” that could let the assistant place orders, book rides, and navigate apps on your behalf.
Gemini’s upcoming capability, spotted by 9to5Google in the Google app 17.4 beta, is internally codenamed bonobo and appears under the Labs feature “Get tasks done with Gemini.”
According to strings uncovered in the APK, the assistant will be able to interact directly with certain apps, performing multi-step actions like ordering food or booking transport without you tapping through menus yourself. This functionality relies on groundwork laid in Android 16 QPR3, which introduces the technical hooks needed for such automation.
Google stresses that Gemini will not be fully autonomous. Users will be able to stop the AI at any time and manually take over, with warnings that “Gemini can make mistakes.”
The company also highlights privacy concerns: screenshots of in-app interactions may be reviewed by trained staff if “Keep Activity” is enabled, and users are advised not to enter sensitive details, such as login or payment information, in Gemini chats. Google explicitly cautions against using screen automation for emergencies or sensitive tasks.
Interestingly, the same beta build also references a “Likeness” feature codenamed wasabi. This appears linked to Android XR’s 3D avatars, currently used in Google Meet calls, with strings suggesting users could access or retake their likeness for prompts. Privacy notes emphasise that “your likeness can only be used by you,” hinting at a controlled rollout of avatar integration alongside Gemini’s automation features.
This is the next step of Google’s push into agentic AI, where assistants move beyond answering questions to completing tasks.
Similar experiments have already appeared in Chrome’s Auto Browse feature, which lets Gemini autonomously navigate web pages. Bringing this to Android could mark a significant shift in how users interact with their devices, effectively turning Gemini into a hands-free operator for everyday tasks.
The thing is, if Gemini’s screen automation reaches public release, it could streamline routine actions like booking a train ticket or ordering groceries. However, the privacy trade-offs and Google’s own warnings mean it will likely launch as an experimental Labs feature first, with limited app support.
As with many AI-driven tools, the balance between convenience and caution will be crucial.
The post Gemini might soon be able to control your phone screen appeared first on Trusted Reviews.
