Kai Aizen

I'm Kai Aizen—security researcher, framework creator, and the mind behind "The Jailbreak Chef."

I came up through the trenches of traditional offensive security. Cloud penetration testing. Web application assessments. And more WordPress security audits than I'd like to admit—digging through plugin code, hunting for SQLi, chasing authentication bypasses in codebases held together with duct tape and prayer. That work gave me six CVEs and taught me how systems actually break. Not in theory. In production.

But my foundation isn't purely technical. I studied political science and psychology—the mechanics of influence, persuasion, and manipulation. I wrote a book about it. For years I thought those fields had nothing to do with security. I was wrong.

When I pivoted to AI security, everything clicked. The same principles that make social engineering work against humans work against language models. Jailbreaking isn't just clever prompting—it's exploiting how systems process trust, context, and authority. Prompt injection isn't a bug class; it's a persuasion problem with technical consequences. The dynamics that let an attacker bypass a help desk can bypass a guardrail.

That insight became my edge.

Today I specialize in adversarial AI: LLM vulnerabilities, agentic AI attack surfaces, and systematic threat modeling. I created the AATMF framework—now on OWASP's 2026 GenAI Security roadmap—to bring structure to a field that desperately needs it. I build tools that automate what I've learned and publish research that bridges offense and defense.

I still break WordPress plugins for fun. Old habits.