11h ago

Researchers Jailbreak Gemini VLM by Replacing Bombs With Bananas in Images

0
Original post

Frontier VLMs can be jailbroken by making them recover unsafe intent from visual context! Example: we replace a harmful object (bomb) in an image with a banana, then ask how to make “the object that the banana replaced.” @GeminiApp complies.

9:26 AM · May 19, 2026 View on X