Google is using user media files to train AI models
Google has begun using media files uploaded by users through search services to enhance artificial intelligence algorithms. This feature is automatically enabled for all users.
According to Engadget, the update includes the use of a wide range of content, such as images, audio, and videos, which users interact with through Google’s search services. For example, photos uploaded for visual search through Google Lens may subsequently be used to train AI models. Similarly, voice search audio recordings may be used in training, as well as files uploaded to Google Translate.
This update applies exclusively to products associated with search systems, while personal files in Google Photos remain untouched.
To disable the data collection feature, users must manually change the settings. To do this, they need to visit the search services history page and uncheck the “Save media” option and also check the personalization settings to ensure that the system does not save any information. Additionally, it is possible to deactivate the display of AI Overview function results.
It is worth noting that Cloudflare plans to change the policy on web crawlers that simultaneously index content for search engines and participate in training artificial intelligence models. From September 2023, the platform will block such crawlers, attempting to return control over content usage to website owners.
Additional:
| To disable data collection | Steps |
|---|---|
| Media storage | Visit the search services history page and uncheck “Save media” |
| Personalization | Ensure the system saves no information on the personalization page |
| AI Overview | Disable the display of results |
