whoami
KAI AIZEN
THE JAILBREAK CHEF
Security Researcher & 5x CVE Holder. Creator of the AATMF and P.R.O.M.P.T frameworks. Specializing in LLM jailbreaking, prompt injection, and adversarial AI.
KAI AIZEN
5
CVE Disclosures
3
Security Frameworks
20+
Published Articles
1
Published Book
Featured Research
jailbreaking GPT original-research
Context Inheritance Exploit: Jailbroken Conversations Don't Die
Discovering how jailbroken states persist across GPT sessions through context inheritance.
January 4, 2025
prompt-injection ChatGPT backdoor
The Custom Instruction Backdoor
Uncovering emergent prompt injection risks through ChatGPT custom instructions.
May 18, 2025
MCP prompt-injection agentic-AI
MCP Security Threat Analysis
Comprehensive security analysis of the Model Context Protocol vulnerabilities.
May 18, 2025
Security Frameworks
AATMF
Adversarial AI Threat Modeling Framework. 14 tactics, 40+ techniques, quantitative risk scoring.
P.R.O.M.P.T
Systematic prompt engineering methodology. Purpose, Results, Obstacles, Mindset, Preferences, Technical.
SEF
Social Engineering Framework. Structured methodology for SE assessments. Coming soon.