I found out last week that Google is using my search history, Gmail, and Google Docs to train its AI models. Not hypothetically — they confirmed it.
The opt-out process exists, but it is buried under six layers of settings and uses language designed to make you second-guess yourself. Classic dark pattern.
As a CS student, this hit different. My Google account has years of academic work, research notes, and project code. All of it potentially feeding into Google's training pipeline.
What Google is actually collecting
Let me break it down:
Search history — Every query you have ever typed. That includes the embarrassing ones, the medical symptoms, the late-night rabbit holes.
Gmail content — Google scans email content for ad targeting. They say they stopped for ads, but AI training is a different bucket.
Google Docs/Sheets — If it is in Google Drive, it might be in the training data. The terms of service are vague enough to allow it.
YouTube watch history — What you watch, how long you watch, what you skip. All signal.
Chrome browsing data — If you use Chrome with sync enabled, your browsing history is fair game.
How to opt out (it is not easy)
Google's opt-out is split across multiple settings pages. Here is the path:
- Go to
myaccount.google.com - Navigate to Data & Privacy
- Find "Web & App Activity" — turn it off
- Find "Ad Personalization" — turn it off
- Go to
myaccount.google.com/data-privacy/web-activity - Turn off "Include Chrome history and activity from sites, apps, and devices that use Google services"
But here is the catch: turning these off does not delete data already collected. It only stops future collection. To delete existing data, you need to go to myactivity.google.com and manually delete everything.
Google makes this deliberately tedious. The delete button is hidden. Bulk deletion requires multiple confirmations. They are betting you will give up halfway through.
Why this matters for students
We are in a unique position. Our academic work, research, and code projects are all stored in Google's ecosystem. Universities use Google Workspace. Most students use Gmail as their primary email.
This means Google has a comprehensive dataset of our intellectual development. Every essay, every coding assignment, every research paper. And they are using it to build products that compete with the very tools we might build ourselves.
It feels like a raw deal.
What I am doing about it
I have started migrating critical work to alternatives:
- ProtonMail for email (encrypted, no scanning)
- Notion or Obsidian for notes (local-first options)
- GitHub for code (not Google's ecosystem)
- DuckDuckGo for search (no tracking)
For AI-assisted work, I have been experimenting with MonkeyCode. The key selling point for me: private deployment. I can run it on my own infrastructure, and my code and prompts never leave my control. No training data extraction, no terms-of-service ambiguity.
As students, we need to be more intentional about where our data goes. The tools we choose today shape the AI landscape of tomorrow.
What is your data privacy setup? Are you still all-in on Google, or have you started migrating? Would love to hear what other students are doing.
Top comments (0)