Mohanakannan

MSc Cyber Forensics & Information Security | CISSP

Cloud & AI Security | Offence & Defense Engineering

Building secure AI systems with an attacker mindset — preventing prompt injection, data leakage, and unsafe behavior.

Prompt Injection Defense RAG Security AI Guardrails Adaptive Defense

View GitHub Explore Projects

👋 About

I am building a hands-on AI Security portfolio focused on attacking and defending real-world AI systems including prompt injection, intent-aware defense, sequential attack detection, adaptive learning, and output sanitization.

👑 What is Double Crown?

Double Crown represents mastery of both sides of AI security:

  • Offensive AI Security — understanding attacks
  • Defensive AI Engineering — building resilience

This enables designing AI systems that are secure against real-world adversarial behavior.

🔐 Core Platform

AI Security Lab

Practical lab for building, attacking, and securing AI systems using layered defenses.

Attack Coverage

  • Prompt injection
  • Role override
  • System prompt extraction
  • Multi-step attacks

Defense Layers

  • Intent-aware detection
  • Sequential defense
  • Adaptive learning
  • Output sanitization
View Project

📊 AI Security Dashboard

Red Team Evaluation Dashboard

Visual tracking of attacks, blocks, leakage attempts, and defense performance.

Measures how well layered AI defenses withstand adversarial prompts.

View Dashboard Code

🧪 Free AI Security Tools

Prompt Injection Detector

Analyze prompts for injection risk and leakage indicators.

Open Tool

🧠 System Focus Areas

Input Defense

Detect malicious prompts before reaching the model.

Decision Intelligence

Risk scoring and adaptive decisions.

Output Protection

Prevent system prompt leakage.

✍️ Latest Posts

📫 Connect