AI Red-Team Academy
AI Red-Team Academy
Learn how modern AI systems really work and how they fail.
From AI fundamentals to red teaming LLMs, RAG pipelines, and agents!
What Is AI Red-Team Academy?
AI Red-Team Academy is a practitioner-focused course that teaches you how AI systems are built, how they are deployed in real organizations, and how attackers exploit weaknesses across the entire AI ecosystem, not just the model.
AI systems are no longer standalone models. They are complex socio-technical systems consisting of: data pipelines, agents, tools, CI/CD, cloud infrastructure, governance, and humans.
Most AI security failures do not come from exotic model hacks. They come from assumptions:
- Assumptions about evaluations and benchmarks
- Assumptions about safety filters
- Assumptions about tools and agents
- Assumptions about humans in the loop
This course teaches you how to break those assumptions.
What You Will Learn
- How AI, ML, and LLMs actually work (no hype, no magic)
- Transformer architecture, tokenization, and prompting mechanics
- How to run LLMs locally and use them in red-team workflows
- Fine-tuning models and working with open-source LLMs
- Vector databases and RAG pipelines
- Building and testing AI agents
- Red-team techniques for LLMs, RAG systems, and agents
- System-level, supply-chain, and human-in-the-loop attacks
- AI-Powered Red Teaming: Accelerating Ethical Hacking with Generative AI
Free vs Member Content
The AI Red-Team Academy follows a transparent model:
🎓 Free on YouTube
https://www.youtube.com/playlist?list=PL-D1lWyrp4YOmiVHvJbJrEn19OvkjZ1E_
Build a rock-solid foundation with our 100% FREE YouTube series covering AI essentials, architectures, pipelines, high-level red team methodology, and tons of practical ‘Red Teaming AI’ lessons. Dive in today and level up your skills without spending a cent!
🔐 Member-Only Content
https://www.patreon.com/cw/jarnobaselier
Ready to go beyond the basics? Our premium content unlocks more advanced red team techniques, in-depth testing methodologies, real-world failure patterns, exclusive course slides, and live workshops. Join as a Patreon supporter today and gain instant access to these high-level resources. Level up your red teaming skills like never before!
Premium Membership Benefits
Knowledge Explorer
- Private Discord Community — Collaborate with fellow practitioners, discuss techniques, share code, and exchange real-world best practices.
- Full Access to Academy Slides — Download and reference all official AI Red-Team Academy presentation decks.
- Exclusive Member-Only Lessons — Access advanced lessons not published on YouTube.
Master Learner
Includes Everything from the “Knowledge Explorer” subscription, plus:
- 🛠 Live Workshops — Participate in live AI Red-Team Academy workshops.
- 📦 Templates & Code — Get access to reusable templates and code as they are released.
AI Red-Team Academy – Full Syllabus
Phase A – Fundamentals
- 1. Basic Terminology (AI, ML, DL, LLM, supervised vs unsupervised, overfitting, generalization) — Learn to speak the language of AI and avoid flawed assumptions later. Free on YouTube
- 2. Public Chatbots and AI Tools — Understand how modern LLM-powered tools work and where their limits are. Free on YouTube
- 3. Neural Network Basics (layers, activations, loss, optimizers) — Understand how learning actually happens inside a model. Free on YouTube
- 4. Transformer Architecture — Learn how attention, embeddings, and tokens power modern LLMs. Free on YouTube
- 5. Tokenizers & Prompting — Learn how models interpret input and how to steer output effectively. Free on YouTube
- 6. How It All Fits Together — Connect all foundational concepts into one coherent mental model. Free on YouTube
- 7. Build Your Own AI Machine — Learn which hardware matters for real-world AI workloads. Free on YouTube
- 8. Python & PyTorch Crash Course — Move from theory to hands-on experimentation with models. Premium Patreon Content
Phase B – Practical Machine Learning
- 9. Train a Small Model — Train a model end-to-end to build deep confidence and understanding. Premium Patreon Content
- 10. HuggingFace Model Hub & Fine-Tuning — Learn how modern AI workflows reuse and adapt existing models. Free on YouTube
- 11. Using Open Models (7B / 13B & Quantization) — Run powerful LLMs locally and understand trade-offs in quality and latency. Free on YouTube
Phase C – RAG Systems
- 12. Vector Indexing (FAISS, Chroma, Milvus) — Embed and search your own knowledge efficiently. Free on YouTube
- 13. RAG Pipeline — Build production-grade, search-powered AI systems. Free on YouTube
- 14. Hands-on: Build Your Own RAG Assistant — Create a private AI assistant that truly knows your data. Premium Patreon Content
Phase D – Safety & Reliability
- 15. Hallucination, Accuracy & Factuality Testing — Measure, test, and control AI behavior to maintain trust and compliance. Free on YouTube
Phase E – AI Agents
- 16. AI Agent Basics — Turn LLMs into systems that take actions, not just generate text. Free on YouTube
- 17. MCP Servers Explained — Understand how tools, data, and models connect — and why this is a key attack surface. Free on YouTube
- 18. Sandbox, Audit & Allowance Flows — Design agents responsibly with full oversight and control. Free on YouTube
- 19. Hands-on: Build a Tool-Using Agent — Build an agent that writes reports or uses tools automatically. Premium Patreon Content
Phase F – Red Teaming AI
- 20. Common Red-Team Techniques (Overview) — Learn how AI systems are commonly attacked and tested. Free on YouTube
- 21. Prompt-Based Attacks on LLMs — Understand jailbreaking, prompt injection, and filter bypasses. Free on YouTube
- 22. Adversarial Attacks on ML Models — Learn evasion and poisoning attacks against classifiers. Premium Patreon Content
- 23. Privacy and Data Attacks — Explore model inversion and membership inference techniques. Free on YouTube
- 24. System-Level & Supply Chain Attacks — Understand how attackers compromise AI pipelines and dependencies. Free on YouTube
- 25. Tool Abuse & Agent Escalation Attacks — Where real-world AI damage happens — beyond prompts alone. Free on YouTube
- 26. Model Confusion & Context Manipulation — Subtle attacks exploiting ambiguity and trust assumptions. Premium Patreon Content
- 27. Evaluation Evasion & Safety Bypass Testing — Learn why models pass tests but fail in production. Free on YouTube
- 28. Human-in-the-Loop Exploitation — Exploit the human component of AI systems — often the weakest link. Free on YouTube
- 29. Agentic AI Red Teaming — Red team systems where humans and AI collaborate as agents. Premium Patreon Content
- 30. Final Conclusion – Lessons Learned — Tie all red team insights together into a practical framework. Free on YouTube
Extras (Members Only)
Additional explanations, deep-dive lessons, live workshops, reusable templates, and code examples are released exclusively for Patreon members.