By design, it only takes a little prompting to hijack the biggest chatBots:• Researcher Dave Kuszmar discovered multiple...

By design, it only takes a little prompting to hijack the biggest chatBots:• Researcher Dave Kuszmar discovered multiple systemic vulnerabilities that let him bypass LLM safety and obtain dangerous instructions.• These exploits worked across nearly all major LLMs revealing an industry-wide security problem.• Kuszmar calls for slowing deployment, increasing transparency, and large-scale research into LLM safety before further integrating these systems into society.https://spectrum.ieee.org/jailbreaking-llms#publicSafety #safety #LLM #privacy #hacking #stochasticParrots #chatBots