Google’s new “Personal Intelligence” feature doesn’t just personalize answers. It quietly reconstructs your life.
For years, trusting Google with our emails, photos, and searches felt like a reasonable trade-off. Convenience in exchange for data. Navigation instead of maps. Search instead of memory.
With Gemini’s new Personal Intelligence feature, that trade just changed — fundamentally.
Once activated, Gemini will be able to draw on your emails, photos, searches, and even your YouTube viewing history.
It’s a powerful feature that improves the relevance of responses, but it’s also a little creepy. We now have a first glimpse of the amount of personal data Google has at its disposal and what the company is capable of doing with it.
This isn’t about access. It’s about inference.
Gemini doesn’t merely read your emails or scan your photo library. It connects dots you never explicitly drew.
In early tests, Google’s AI was able to infer !
Pranav Dixit from Business Insider tested Personal Intelligence. He asked for suggestions of places to visit for his parents in the San Francisco Bay Area, and the AI suggested museums and gardens, as his parents had already been hiking and on excursions in redwood forests.
The chatbot indicated that it had drawn on photos, family emails, a parking reservation in Gmail and his search history (including “easy hikes for seniors”). It should be noted, in case there is any doubt, that this data came from the journalist’s account, not his parents’.
Google’s chatbot is now able to draw on all Google services to exploit your data. © Image generated with ChatGPT
Based on photos, Gemini was able to tell him his car’s license plate number. Thanks to his emails, the chatbot could tell him when he needed to renew his insurance. When planning a trip, the AI had already taken into account the fact that they would be traveling with a young child, as Gemini already had this kind of data.
Entrusting your data to these services seemed reasonable a few years ago, before the era of AI. At the time, Google couldn’t know what was in our photos unless employees viewed them directly. Now, that’s no longer necessary. Gemini can analyze a large part of our digital lives, and most certainly discover details that we thought were private, or that we were completely unaware of about ourselves.
Based on photos, Gemini was able to tell him his car’s license plate number. Thanks to his emails, the chatbot could tell him when he needed to renew his insurance. When planning a trip, the AI had already taken into account the fact that they would be traveling with a young child, as Gemini already had this kind of data.
Entrusting your data to these services seemed reasonable a few years ago, before the era of AI. At the time, Google couldn’t know what was in our photos unless employees viewed them directly. Now, that’s no longer necessary. Gemini can analyze a large part of our digital lives, and most certainly discover details that we thought were private, or that we were completely unaware of about ourselves.
None of this came from the parents’ accounts. It came from patterns, habits, reservations, images, and past searches. From context.
The shift most people are missing
The uncomfortable truth is this: Google no longer needs to “look” at your life to understand it.
Where older systems required explicit input — forms, queries, declarations — Gemini operates on implicit signals. It doesn’t ask who you are. It infers who you’ve become.
That’s a qualitative leap. Not a feature update. A new cognitive layer.
Privacy is the wrong framing
Google insists Gemini is not directly trained on your personal data. Technically, that may be true.
But that reassurance misses the point entirely.
The risk is no longer raw data misuse. The risk is behavioral reconstruction. An AI that knows where to find what it needs — and when to use it — without ever storing it in one place.
Control doesn’t disappear. It becomes abstract.
Why Google holds an unmatched advantage?
No other AI company comes close to this position.
Google doesn’t just power a chatbot. It sits across email, search, maps, photos, video, mobile operating systems, cars, watches, TVs, and browsers.
Gemini isn’t learning about you in isolation. It’s learning across your entire digital surface area.
That makes “personal intelligence” less a feature — and more a mirror.
When anticipation replaces choice
The real question is not whether Gemini can help you. It clearly can.
The question is what happens when systems begin to anticipate needs, nudge decisions, and filter options before you consciously engage.
When assistance becomes pre-emptive, autonomy quietly erodes.
This is the next phase of AI power
We are moving past assistants that respond. Into systems that model.
Model your routines. Your constraints. Your preferences. Your future behavior.
And once an AI understands you better than you articulate yourself, the debate is no longer about data.
It’s about agency.
Personal Intelligence may feel helpful today. But it also marks the moment AI stopped waiting for us to explain who we are.
