Google has quietly updated its privacy settings, allowing the tech giant to store more personal data—including images, files, and audio and video recordings—to train and improve its artificial intelligence (AI) models. Notably, this feature is enabled by default under the guise of giving users more control over their search history and personalized recommendations. Unless users proactively opt out, any media uploaded to Google’s Search services is now being channeled into its AI training pipeline.
Diễn biến chi tiết
According to TechCrunch, this change stemmed from an under-the-radar update to Google's Search services privacy settings, which was announced via a customer email in June 2026. The update introduced two new settings: "Search Services History" and "Personalized Recommendations." These new rules do not just apply to the traditional Google Search engine but also extend across a suite of other search-related services, including Google Maps, Google Shopping, Flights, Hotels, Google Translate, and Google News. This means that users' daily interactions across these platforms now fall within the scope of data collection.
Phân tích kỹ thuật & Công nghệ
On the technical side, the new collection mechanism targets multimodal processing tasks directly. For instance, when a user snaps a photo using Google Lens for visual search, that image may be saved for AI training. Similarly, audio recordings from voice inputs on the new "Search Live" feature in the Google app or pronunciation practices on Google Translate can also be captured. In its support documentation, Google confirms that it uses this history to develop and improve its services, including training generative AI models with the assistance of human reviewers. While some storage is temporary for operational purposes, Google’s official language indicates that saved media can be retained long-term specifically for AI development.
Ý kiến chuyên gia & Nhận định
Security experts note that Google's move reflects a broader industry shift: harvesting user data by any means necessary to upgrade AI capabilities. Instead of relying solely on web-scraped public data, tech giants are increasingly mining user-generated and user-uploaded media from within their proprietary ecosystems. Meta is another prominent example, having faced scrutiny for training AI on users' unshared camera roll photos and content captured by Ray-Ban Meta smart glasses. This trend raises deep privacy concerns as the boundary between personal utility data and machine learning training resources continues to blur.
Tác động & Tương lai
For Vietnamese users, who rely heavily on Google's ecosystem from navigation to translation, reviewing privacy settings has become essential. Fortunately, Google still provides an opt-out mechanism. Users can navigate to their Google Activity controls to uncheck the "Save Media" box independently or disable the entire "Search Services History" setting, as well as configure auto-delete schedules for 3, 18, or 36 months to safeguard their personal privacy in the current era of aggressive AI data harvesting.