If you use Google’s search services, your uploaded photos, voice recordings, and other media may now be feeding the company’s artificial intelligence models. A little-noticed update to Google’s privacy settings, rolled out in June, automatically opted users into allowing their saved media to be used for training AI. This change applies across Google Search, Maps, Shopping, Flights, Hotels, Translate, and News—any service where you can upload or record content.
Previously, Google used your search history and web activity to personalize ads and recommendations, but the new setting expands data collection to include “images, files, and audio and video recordings.” The company’s help documentation explicitly states: “Like your Search Services History, your saved media is also used to develop and improve Google services and technologies, including AI models and safety measures.” This means that snapshots taken with Google Lens, voice queries from Search Live, or audio recorded while practicing pronunciation in Google Translate can all be retained and repurposed for AI training.
What Changed and Why It Matters
The update split Google’s previous “Web & App Activity” into two separate settings: the original Web & App Activity and a new “Search Services History” control. By default, the new setting is turned on, and it includes a sub-option to “Save Media.” Many users were unaware of the change because they only received a single email notification, which many dismissed as routine. This shift reflects a broader industry trend: instead of relying solely on public web data, tech giants are increasingly harvesting user-uploaded content to train their generative AI models. Meta, for instance, trains its AI on user photos and even content from its smart glasses. Google’s move puts it in the same camp, raising privacy concerns even as the company promises human review and safety measures.
The data retention period is configurable—users can choose to have saved media automatically deleted after 3, 18, or 36 months. However, the media collected before you opt out may already be in use for AI training. Google says some storage is temporary and necessary for the product to function, but the company’s language makes clear that saved media can be used specifically to improve AI models, not just for personalization.
How to Opt Out and Regain Control
Fortunately, you can reverse the setting. The primary control is on the Search Services History page. There, you can uncheck the “Save Media” box independently from the “Search Services History” box. If you uncheck both, Google will stop saving new media and your search activity for AI training and personalization.
Additionally, you can access the Search Services Personalization page to limit how your data is used. From that page, you can also navigate to other privacy settings like Web & App Activity, Timeline, and YouTube History to further restrict data collection. A crucial nuance: before this update, changing your Web & App Activity settings would also affect your search data. Now, because the two are separated, you must adjust both sets of controls to fully opt out of having your media and search history used for AI training. If you only modify Web & App Activity, your search data and media remain subject to the new default.
Privacy advocates recommend a review of all Google privacy settings at least once a year. The company’s “My Activity” dashboard (myactivity.google.com) provides a centralized overview of what data is stored and enables you to delete specific items or entire categories. For those concerned about AI training, immediately unchecking “Save Media” is the most direct step.
Implications for Users and the Industry
Google’s decision to silently expand data collection underscores a growing tension between AI development and user privacy. The company argues that user data helps create safer, more accurate AI models—for example, detecting harmful content or improving voice recognition for diverse accents. Critics, however, note that opt-out models place the burden on users to constantly monitor settings, while the company profits from the AI improvements. The European Union’s GDPR and similar laws in California require clearer consent, but Google’s email notification may not meet the legal standard of “unambiguous consent” in some jurisdictions. Several privacy groups have already filed complaints with data protection authorities.
For everyday users, the practical impact varies. If you rarely use Google Lens or voice search, the risk may be low. But for power users who upload many photos, use Translate frequently, or rely on Google Maps’ location history, the setting change could mean a large amount of personal media being used for AI training—including images of your home, family, or sensitive documents.
The broader industry shift is unmistakable. As companies like Google, Meta, and OpenAI compete to improve their AI, user-generated content becomes an increasingly valuable resource. Without stronger regulatory frameworks, the default will remain “collect now, ask later.” Until then, the onus is on individuals to navigate ever-changing privacy settings. Google’s update is a reminder that even trusted services can repurpose your data in ways you may not expect—and that opt-out steps are not always obvious.
To stay protected, regularly check your Google Account settings, disable media saving, and consider using alternative services for sensitive tasks. The tools are there, but they require proactive effort.
Source:TechCrunch News