Gbhackers Google Gemini Safety Filter Bypass Exposes Blocked AI Images
Article Content
- •Google Gemini's safety filter bypass allowed access to blocked AI images via download links.
- •The flaw was disclosed by Fog Security and initially reported to Google in June 2026.
- •No CVEs have been assigned, and Google has not confirmed the root cause of the issue.
Research from Fog Security revealed that Google Gemini allowed users to download AI-generated images that its safety filters had rejected. This bypass did not require a jailbreak; instead, it exploited a flaw in the download mechanism, which sometimes delivered a rejected image while displaying an allowed one in the conversation. The issue was disclosed on November 5, 2026, with initial reports dating back to June 25, 2026, when it was submitted to Google’s Vulnerability Reward Program. Google initially deemed the report out of scope but later resumed analysis after clarification. By October 22, 2026, the issue appeared to be resolved, with no further reproduction of the flaw. However, concerns about potential state-management flaws remain unconfirmed. No CVEs have been assigned to this issue as of now.
Ask AI about this cluster
Answers cite the sources they use
Timeline
More articles in this cluster (3)
Following this threat?
Track Fog Security in your own feed — you're alerted when they show up in new reporting, leak sites or exploitation.
Free account · no card needed
Common questions
What images are affected by this bypass?
Has Google issued a fix for this issue?
Is this vulnerability being actively exploited?
Continue Reading
Critical RCE Vulnerability in Confection WordPress Plugin (CVE-2026-115252) A critical vulnerability, CVE-2026-115252, has been identified in the Confection WordPress Cable plugin, affecting versions 0.4.0 to 0.4.7. This flaw allows unauthenticated attackers to execute arbitrary PHP code on the server by writing unmodified settings into a PHP file without proper checks. The vulnerability has…