Hacking internal AI chatbots with ASCII art is a security team’s worst nightmare

While LLMs excel at semantic interpretation, their ability to interpret complex spatial and visual recognition differences is limited. Gaps in these two areas are why jailbreak attacks launched with ASCII art succeed.
While LLMs excel at semantic interpretation, their ability to interpret complex spatial and visual recognition differences is limited. Gaps in these two areas are why jailbreak attacks launched with ASCII art succeed.Read More