Google updates privacy settings to train AI on user media

Google has updated its privacy settings to allow the company to store user media to improve its artificial intelligence models. Under the update, which Google announced in June, users are opted into this AI training on their media by default. However, users can manually opt out of the data collection.

The update applies to Google Search and several other search services, including:

• Maps
• Shopping
• Flights
• Hotels
• Translate
• News

The stored data includes “images, files, and audio and video recordings,” according to Google. In an email to customers, the company stated: “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.” Here, Search Services History refers to a Google-specific privacy setting that manages how user search activity is saved. According to Google’s help documentation, the company uses this history to provide, develop, and improve its services, which includes training generative AI models, and to protect Google, its users, and the public with the assistance of human reviewers.

The policy change directly impacts features like Google Lens, a visual search tool, where captured images may be saved for AI training. Similarly, audio recordings from Search Live—a voice search feature—or other voice searches could be saved for AI training.

To manage these settings, users can configure how often their saved data is automatically deleted. Google offers data deletion intervals of three months, 18 months, or 36 months. Previously, Google allowed users to configure what historical search data was saved via Web & App Activity, a Google-specific data retention setting. Following the update, this has been separated into Web & App Activity and the new Search data setting. Because these are now separate, changing the Web & App Activity settings will no longer impact the storage of Google Search services data.

This change reflects a wider industry shift in how technology companies source training data. Rather than relying solely on scraping public websites, companies are increasingly using content that users upload or create. For example, Meta is also training its AI on users’ images and media.

Why it matters

This shift represents a broader industry trend where tech giants are increasingly leveraging user-generated content to train AI models, moving beyond public web scraping to proprietary data collection.