Google Stores Uploaded Media From Search Interactions to Train AI Models

*Google now keeps images and other media from Search interactions in user history for AI training unless users change their settings.*

Google made a quiet update to Search history that retains media uploads, including images submitted for reverse image searches. The stored files become part of the data used to train the company’s AI models.

The change affects any user who has Search history enabled. Previously, media uploaded during a session stayed tied only to that session. The new policy keeps the files unless the user turns off the relevant training option.

No public numbers have been released on how many accounts are affected or how much additional media is now retained. The company has not disclosed the exact retention period or downstream uses beyond model training.

Reactions

No statements from competing search providers or privacy groups appear in the reporting on this update. Google has not published an official blog post detailing the scope of the change.

Why it matters

Users who rely on visual search features now face a default that funnels their uploads into training data. Anyone who wants to avoid that outcome must locate and disable the setting before the next upload. The shift shows how routine interactions with search are being turned into raw material for larger models without requiring separate consent at the moment of upload.

---

Sources:

• Wired

{
  "excerpt": "Google now stores media uploads from Search interactions for AI training unless users opt out through their history settings.",
  "suggestedSection": "ai",
  "suggestedTags": ["google", "ai-training", "privacy"],
  "imagePrompt": "Abstract composition of scattered translucent data fragments floating above a dark reflective surface, suggesting stored media and model inputs. Muted color palette, cinematic lighting, 16:9."
}