Google announced in an email that starting July 7, 2025, your Android will allow Gemini AI to interact directly with core apps, regardless of whether you've already turned off the connection. With the instructions for turning it off being sparse and ambiguous, users are wondering if the AI will be reading or even sending messages without their knowledge.
Imagine your phone becoming an autonomous agent: opening your messaging apps without asking, reading your conversations, and “helping” you transcribe your responses. Sound like something out of a science fiction movie? Gemini AI and Google wants this to become your reality next week. But when someone sneaks in a feature without clear instructions on how to turn it off, it begs the question – is your private digital living space really still yours?
What does “help” mean to Gemini AI?
Google promises in the announcement that Gemini will “help you use the Phone, Messages, WhatsApp, and Tools apps” starting July 7. But the word “help” can mean a lot:
- Automated reading and suggestion: AI could scan your messages and offer you suggested responses or even text translation in real time.
- Automatic execution of commands: Imagine Alexa making the call for you – but now Gemini does it.
- Context capture and profiling: AI could analyze conversation patterns, contact list, and even call history to customize its suggestions.
However, because Google does not explain where and how these "helpful" options can be turned off, users remain uncertain about whether it is possible to keep their communications private.
Where to find the off button?
The notice only states: “If you don’t want to use these features, you can disable them in the app settings page.” However, since Android’s permissions and activity management menu is notoriously complex, here’s a summary of the expected steps:
Open the Gemini app (mobile or web).
- Click on your profile in the top corner and select “Gemini Apps Activity”.
- Disable it “Auto-save activities” for selected applications.
- If you wish, delete the data already collected – but be careful: Google may store interactions for up to 72 hours to ensure the smooth operation of the service.
- While these steps sound promising, the question remains whether this will actually disable all “assistant” access to your contacts, messages, and calls. Google is keeping quiet for now.
Why is privacy concern not exaggerated?
- Unclear definition of permissions: The email text does not explain whether “helping” includes automatically sending or modifying messages.
- Hard-to-reach setting: Although the goodies are hidden in Gemini Apps Activity, there is no guarantee that some APIs will only work if this feature is completely disabled.
- Obstacle in case of possible shutdown: Google Assistant has often asked to re-enable certain permissions in the past, even though they were once disabled.
Due to these aspects, it is obvious that Google is targeting as many users as possible, while at the same time leaving the door open for "proactive" intervention just in case.
What awaits us after July 7th?
Next week will reveal how consistently Google is implementing its ambitions. Users who previously mistakenly believed they had turned off all Gemini AI interactions may be unpleasantly surprised to see their phones behaving strangely. On the other hand, those who dared to trust Google’s repeated explanations may find the AI providing an incredibly useful experience—from automatic call sorting to intelligent message suggestions.
Conclusion: Gemini AI
If you’re a privacy enthusiast (or just worried about what AI can do with your messages and calls), now’s the time to navigate your settings. Before July 7, manually review and turn off anything called “Gemini Apps Activity,” and delete your interaction history if necessary. If Google wants to keep you safe, it should clearly explain this miraculous on-off switch—not hide it in fancy but vague terms. And while you’re at it, check out what’s going on in your phone’s other permissions. You can never have too much of your privacy.
