When you snap a photo with Google Lens to identify a plant or record a voice note to translate a phrase, you are likely thinking about the immediate result. You aren't thinking about the data center where that image or audio file is being archived. But according to a recent, under-the-radar update to Google’s privacy policy, you should be.

Google has begun using user-uploaded media—including images, files, and audio recordings—to train its generative AI models. The change, which rolled out via a quiet customer email in June, effectively opts users into this data-harvesting program by default. It is a significant shift in how the company treats the content you create while using its suite of services, from Maps and Shopping to Translate and News.

The New Default: Your Data as Training Fuel

The update introduces two primary settings: "Search Services History" and "Personalized Recommendations." While these sound like standard features designed to improve your user experience, they serve a dual purpose. Google’s own help documentation now explicitly states that saved media is used to "develop and improve Google services and technologies, including AI models and safety measures."

This isn't just about text-based search queries anymore. The scope of this data collection is broad:

  • Visual Data: Images uploaded to Google Lens are now fair game for training.
  • Audio Recordings: Voice searches and audio captured during translation practice are being stored.
  • Cross-Service Integration: This applies across the entire Google ecosystem, meaning your activity in Maps, Flights, and Hotels is now linked to these training pipelines.

Why This Matters Now

The industry is currently in a desperate race for high-quality training data. As the pool of publicly available, high-quality web content begins to dry up, tech giants are increasingly turning to the "walled gardens" of their own platforms. By capturing the media users generate in their daily lives, companies like Google and Meta are securing a proprietary pipeline of data that competitors cannot access.

For the user, this creates a new layer of complexity. Previously, managing your privacy meant toggling "Web & App Activity" settings. Now, Google has decoupled "Search Services History" from that master switch. If you previously opted out of data retention, you may find yourself opted back in under these new, specific service categories.

How to Take Back Control

You do not have to accept this as the cost of using Google’s tools. You can manually disable these features, though you have to navigate to specific sub-menus to do so.

  1. Access Search Services History: Navigate to your Google account’s data and privacy settings. Look for the "Search Services History" section.
  2. Uncheck "Save Media": Within this menu, you can specifically uncheck the box that allows Google to store your images, files, and audio for AI training without disabling your entire search history.
  3. Set Auto-Delete: If you prefer to keep some history for convenience, configure the auto-delete setting to three months. This ensures that your data is purged from Google’s servers on a rolling basis, limiting the window of time it can be used for model training.

Key Takeaways

  • Default Opt-In: Google is now using your personal media—including photos and voice recordings—to train its AI models by default.
  • Decoupled Settings: Your old "Web & App Activity" privacy settings no longer cover these new search services; you must opt out of them separately.
  • Granular Control: You can disable the "Save Media" feature while keeping other search history functions active, allowing for a balance between utility and privacy.

What This Means for Users

This update is a reminder that in the era of generative AI, the definition of "personal data" has expanded. It is no longer just your search history or your location; it is the photos you take and the voice notes you record. While Google argues this is necessary to improve safety and model performance, the burden of privacy has shifted entirely to the user.

If you value your digital footprint, the next step is to audit your account settings. The default state of the internet is now "data collection," and it will remain that way until you manually flip the switch.