ChatGPT safety systems can be bypassed to get weapons instructions

NBC News found that OpenAI’s models repeatedly provided answers on making chemical and biological weapons....

Redirecting to full article...