Google Gemini Safety Filter Bypass Allows Blocked Image Downloads
Original title: Google Gemini Safety Filter Bypass Exposes Blocked AI Images Through Download Links
Gemini exposed rejected images via download links after safety filter refusal.
Fog Security researcher found Google Gemini web app exposed images rejected by safety filters through separate download paths. The download could retrieve previously blocked artifacts while the chat displayed approved images. Researchers reported the issue to Google after reproducing it in conversations.
Unconfirmed: Root cause is a hypothesis about state-management flaw in image generation and safety interruptions
- Separate paths for image viewing and download
- Download retrieved blocked artifacts
- Behavior changed after maintenance
- No privilege escalation or cross-user exposure
Impact: Access to blocked AI-generated images
- 1.Bind downloads to displayed approved images
- 2.Remove artifacts from refused generations
From the original article
3 min read at gbhackers.com
Google Gemini exposed images that its safety filters had rejected through download links, revealing an infrastructure weakness rather than a conventional model jailbreak. Fog Security researcher Jason Kao disclosed the findings on November 5, 2026, describing how Gemini could refuse an image request while still delivering the blocked artifact. Retesting on October 22 found changed […]
By Divya · gbhackers.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.