From the episode description
Welcome to ResearchNode’s AI Security Testing series.
This introduction explains the course roadmap, what you’ll learn, and how the weekly lessons fit together. We’ll start by mapping an AI application, then explore how weaknesses in instructions, permissions, data and deployment can affect its behavior.
The aim is practical: understand the issue, design a focused test, collect evidence, and check whether a fix works.
THE 12-LESSON ROADMAP
01 — How to test an AI system
02 — Prompt injection
03 — MCP and tool permissions
04 — Unsafe AI output
05 — RAG poisoning
06 — Embedding privacy
07 — Memory poisoning
08 — Agent impersonation
09 — Training data poisoning
10 — Malicious model files
11 — AI deployment security
12 — Security retesting
Each lesson uses clear diagrams, original examples and code where it helps explain the test. Mock scenarios and measured results are clearly distinguished, so you can understand what the evidence actually shows.
One new main lesson is planned each week, with videos around eight minutes.
OPTIONAL HANDS-ON PRACTICE
You can follow the entire series without a lab account or installing software. If you want to practice, each lesson’s description points to the relevant P2501 lab and manual.
P2501: https://p2501.ai
The methods apply beyond P2501 to other AI systems you’re authorized to assess. ResearchNode’s AI Foundations series provides useful background if these concepts are new to you.
Continue with Lesson 01: How to Test an AI System.
#AISecurity #SecurityTesting #ResearchNode
Descriptions and chapter markers are supplied by the channel. They are not a transcript or an independent check of every claim in the video.
