Google Trains AI on Your Media by Default
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Google updated its Search services privacy settings in June, introducing "Search Services History" and "Personalized Recommendations" settings that store user media — including images, audio, and video recordings — for AI training by default
- The policy change extends beyond Google Search to Maps, Shopping, Flights, Hotels, Translate, and News, meaning photos snapped via Google Lens and voice recordings from Search Live and Translate can all be saved for model development
- Google confirmed in a customer email that saved media is "used to develop and improve Google services and technologies, including AI models and safety measures," and its help documentation notes human reviewers may also access the data
- Users can opt out by unchecking the "Save Media" box on the Search Services History page, with options to set auto-deletion at 3, 18, or 36 months
- The update split Google's prior "Web & App Activity" controls into a separate setting, meaning existing opt-out configurations no longer cover search service data
- Meta is cited as another consumer tech company training its AI on user-uploaded images, media, and content recorded by its AI glasses
Why it matters: Google's opt-in-by-default approach means any user who never reviewed their privacy settings is now contributing personal media to AI model training, and splitting Web & App Activity into a separate setting means prior opt-outs no longer cover search data — a material expansion of the company's training pipeline that many users will not realize happened.
