Google's Gemini: Screen Automation Sparks Privacy Fears

Google is significantly advancing its AI assistant, Gemini, by integrating agentic capabilities that allow it to independently control applications on Android devices. This new "screen automation" feature, reportedly in development, aims to let Gemini perform tasks within certain apps, initially for services like ride-sharing or ordering.

However, this development raises considerable privacy concerns. Code suggests that Gemini's actions are logged, reviewed by human staff, and that users should avoid entering sensitive information. The AI's potential for errors is acknowledged, with users ultimately responsible for its actions. Manual override options will be available, but the author questions the necessity and safety of such automation.

This move reflects a broader trend towards AI deeply embedding into smartphones, potentially enabling them to manage most functions. Despite personal reservations about AI over-reliance and privacy, the author acknowledges this technological progression appears inevitable given user adoption rates. They plan to disable such features where possible.