The European Union has directed Google to grant competing AI assistants equivalent access to sensitive Android resources, including the microphone, camera, and screen content. The decision aims to ensure that third-party services receive the same capabilities currently available to Gemini.
The measures encompass eleven core system functions such as voice-command activation, home-button invocation, background execution, and direct access to on-device AI models. Alternative assistants will be able to use these resources without being set as the default service.
Environmental data access will provide real-time streams from the microphone, camera, screen, and speakers. All releases must follow the same consent and notification standards that apply to Google’s own services.
AI tools will also be permitted to interact with other applications, send messages, schedule appointments, adjust device settings, and execute multi-step tasks. Screen automation will enable assistants to open applications inside a virtual window and simulate taps and typing while the user continues other activities.
The majority of these changes are required to ship in Android 18 by 1 August 2027. Simultaneous activation of multiple assistants via voice keywords will be implemented in Android 19, with a final deadline of 1 August 2028.
Every function will remain subject to explicit user authorization. For particularly sensitive operations such as screen automation, system integration, and centralized app-data queries, Google may impose objective security and certification requirements.


