Module 1: AI Architecture & Agentic Foundations
- Trace the development of AI from Turing's test to modern agentic systems
- Demystify ML, deep neural networks, transformers, and LLMs
- Master agentic AI design patterns: Cognitive Loop, Planner-Executor-Verifier, multi-agent orchestration
- Identify the AI Security Ecosystem attack surface across compute, data, model, and agent pipeline layers
Module 2: Generative AI for SecOps and Risk Management
- Deploy RAG pipelines integrating live threat intelligence with chunk provenance validation
- Build AI-powered security operations workflows including incident reporting chatbots
- Establish a strong foundation in AI security risk management (CIA Triad, CVE, GenAI-specific risks, DLP)
- Apply adaptive authentication and data protection patterns to AI system deployments
Module 3: Hacking AI Agents – Adversarial Techniques
- Identify OWASP ML Security Top Ten and OWASP LLM Top Ten risks
- Execute the full prompt injection taxonomy: direct, indirect, chained, multi-language, refusal suppression
- Master jailbreaking (DAN), prompt leaking, and agent hijacking via crafted inputs
- Apply MITRE ATLAS and NIST AML taxonomy; execute AI Red Teaming methodology
- Understand GenAI social engineering, deepfake attacks, and the AI offensive toolkit
Module 4: Exploiting the AI Attack Surface
- Conduct training data poisoning, model extraction, and membership inference attacks
- Execute token inference side-channel attacks, hallucination exploits, and payload splitting
- Perform AI-assisted memory forensics using Volatility 3 to detect advanced threats
- Map all attacks to the NIST AI 100-2 taxonomy and MITRE ATLAS matrix
Module 5: Defending with Agents-Autonomous SecOps
- Build autonomous multi-agent threat detection and response workflows with human-on-the-loop oversight
- Integrate AI-based IDS, SOAR playbooks, and threat intelligence into agentic SecOps pipelines
- Deploy multi-agent systems using LangChain/CrewAI with Kafka and Redis/Celery for agent infrastructure
- Augment SIEM and SOAR with GenAI: NLP threat queries, playbook generation, AI-assisted triage
Module 6: AI Governance & Zero Trust for Agents
- Apply NIST AI RMF AI 600-1, OWASP LLM Governance Checklist, and regulatory frameworks to AI agent deployments
- Implement Zero Trust patterns for generative AI and agentic systems
- Deploy a role-aligned AI security agent with signed audit logging as the Zero Trust exit criterion
- Understand quantum computing implications and advanced persistent AI threats for future readiness
