This post is also available in:
עברית (Hebrew)
Android’s digital assistant, Gemini AI, is now extending its reach further into the operating system, automatically integrating with apps such as WhatsApp, even for users who previously restricted such access.
Starting July 7, a quiet system update has granted Gemini default permission to interact with select third-party applications, as part of a broader effort to embed AI more deeply into Android’s ecosystem. While positioned as a user-friendly upgrade, the change has triggered renewed concern from digital rights advocates, particularly over how user data is collected and stored.
Users who had earlier opted out of app-level access may find those settings overridden. According to details circulating in privacy forums and tech news outlets such as Android Police, the AI assistant can still hold onto user data for up to 72 hours—even if data collection features are turned off. Additionally, Google has acknowledged that human reviewers may review data collected through the assistant to improve system performance.
What remains unclear is how to fully opt out. The system does not offer a straightforward method to completely remove Gemini from a device. Instead, users are given partial control over data sharing and app access.
To reduce the assistant’s access, Android users can:
- Disable the Gemini app via the system settings. This halts its activity but doesn’t remove it entirely.
- Limit app integration by turning off access to individual apps within the Gemini settings menu.
- Pause activity tracking, which stops prompt history from being used to improve the AI, though short-term data storage still applies.
As AI becomes more central to smartphone operating systems, the tension between convenience and user control is likely to intensify. For now, Android users concerned about data exposure will need to navigate multiple layers of settings to keep their personal information out of Gemini’s hands.
