AISlashdot1 day ago
How a Seemingly Harmless Image Can Jailbreak Vision-Language AI Models
1 min read
Florida International University researchers developed JaiLIP, a technique using subtle image modifications to bypass AI safety guardrails. The attack works on images appearing normal to humans. Tests against BLIP-2 showed manipulated images significantly increased harmful responses, outperforming traditional prompt-based jailbreaks.
Level
Hype check
Tap to vote and see what everyone thinks.
