Google Uses Uploaded Images and Media to Train AI Unless Users Opt Out
Google now saves images and media uploaded during search interactions to train AI models by default, unless users disable the setting. This excludes Google Photos content and retains data for up to four years. Users can opt out via their account settings.
הידיעה ב-6 שורות · תקציר של baba
- Google saves uploaded images and media during search to improve AI by default.
- Saved media includes Google Lens images, voice search audio, and files for Translate.
- Google Photos content is excluded from this data collection.
- Users can disable media saving in their Google account settings.
- Previously saved media may still be used for AI training up to four years.
- The move reflects growing competition in AI development among tech companies.
Google has introduced a new default setting that allows the company to save images, audio recordings, videos, and other files uploaded during the use of its search services. This data is then used to develop and enhance Google's artificial intelligence models. The change affects media uploaded through search interactions, such as images submitted to Google Lens, voice recordings from voice search, and files uploaded to Google Translate, but does not include content stored in Google Photos.
According to Google, the saved media helps improve user experience by enabling features like returning to previous searches and supports the advancement of technologies powering services such as Lens, Translate, AI Mode, and voice search. When media is used for AI training, it is detached from the user's account and retained for up to four years.
Users who do not want their uploaded content to be used for AI training can disable this feature manually by going to the "Search Services History Settings" in their Google account and unchecking the "Save Media" option. However, turning off this setting only stops new media from being saved; previously saved content may still be used for technology improvements and retained for up to four years if already integrated into AI training.
This update comes amid intensifying competition in the AI sector, where major tech companies seek diverse data sources beyond text, including images, audio, and video, to enhance their AI models.
אותו אירוע, כפי שכל מערכת דיווחה עליו בנפרד. פתחו כמה דיווחים כדי להשוות מה כל מערכת מדגישה, ומה היא משאירה בחוץ.
מרכז 1חרדית 1
