Researchers jailbreak AI chatbots with ASCII art -- ArtPrompt bypasses safety measures to unlock malicious queries

www.tomshardware.com Researchers jailbreak AI chatbots with ASCII art -- ArtPrompt bypasses safety measures to unlock malicious queries

ArtPrompt bypassed safety measures in ChatGPT, Gemini, Claude, and Llama2.

1 crossposts

Technology @lemmy.world shish_mish @lemmy.world 9mo ago

Researchers jailbreak AI chatbots with ASCII art -- ArtPrompt bypasses safety measures to unlock malicious queries

www.tomshardware.com /tech-industry/artificial-intelligence/researchers-jailbreak-ai-chatbots-with-ascii-art-artprompt-bypasses-safety-measures-to-unlock-malicious-queries