AI Won’t Tell You How to Build a Bomb—Unless You Say It’s a ‘B0mB’

Anthropic’s Best-of-N jailbreak technique proves how introducing random characters in a prompt is often enough to successfully bypass AI restrictions.
{Categories} _Category: Inspiration{/Categories}
{URL}https://decrypt.co/298397/anthropic-jailbreak-bypass-ai-restrictions-censorship{/URL}
{Author}Jose Antonio Lanz{/Author}
{Image}https://img.decrypt.co/insecure/rs:fill:1024:512:1:0/plain/https://cdn.decrypt.co/wp-content/uploads/2024/12/9Yzc0liGR-iVmKwK55tNhQ-gID_7.jpg@png{/Image}
{Keywords}Artificial Intelligence{/Keywords}
{Source}Inspiration{/Source}
{Thumb}https://cdn.decrypt.co/wp-content/uploads/2024/12/9Yzc0liGR-iVmKwK55tNhQ-gID_7.jpg{/Thumb}