Overview
An individual who has gained notoriety for consistently finding and exploiting vulnerabilities in AI models, often on their launch days. He uses prompt engineering techniques to bypass safety restrictions and elicit unintended responses.
Role in this knowledge base
The central figure demonstrating AI vulnerabilities, whose actions prompt the discussion and analysis within the video.
Key facts
- Has reportedly jailbroken major AI models like Grok, Kimmy K, and Claude Opus on or shortly after their release dates.