PromptExploit.
AI FUNDAMENTALS / ROOM 02

AI Security Threats.

Learn what can go wrong in an AI system, how to recognise the problem, and where defenses belong.

Read the room notes
AI security risks across data, a model, and connected tools.SMALL IDEAS. CLEAR CONNECTIONS.

Task by task

  1. 01

    Introduction

    Start thinking about AI security: what needs protection, what can go wrong, and how to ask useful questions.

  2. 02

    Vulnerabilities in AI Models

    Understand prompt injection, poisoned data, model theft, privacy leakage, and drift through simple examples and practical defenses.

  3. 03

    AI-Enhanced Attacks

    See how attackers use AI for malicious software, convincing messages, and fake voices, then practise checking the request behind the disguise.

  4. 04

    Defensive AI

    Learn how AI can help investigate security events, read logs, and summarize evidence while people check the facts and control important actions.

  5. 05

    Securing AI

    Protect an AI system with clear permissions, careful data handling, testing, monitoring, and a plan for mistakes.

  6. 06

    Practical

    The ten AI Security Analyst Orientation questions, in order, with answers and simple-English explanations.

  7. 07

    Conclusion

    Connect AI weaknesses, AI-assisted attacks, defensive uses, and secure design with a plain-English recap and revision questions.

Open the original TryHackMe room ↗
THE STUDY RECORD

Learning made visible.

This collection covers all eight tasks in The Building Blocks of AI and all seven tasks in AI Security Threats. Each note brings together plain-English explanations, examples, and sources.

First documented on . Expanded on .

View the earlier study screenshots

These screenshots accompanied the first study entry. They show The Building Blocks of AI and tasks 1–2 of AI Security Threats.

Supplied TryHackMe screenshot: The Building Blocks of AI shows room completed 100 percent and green completion checks on tasks 1 through 8.
The Building Blocks of AI · all eight tasks. Open the image for a larger view.
Supplied progress screenshot with green completion checks on Task 1 Introduction and Task 2 Vulnerabilities in AI Models.
AI Security Threats · tasks 1 and 2 · first study entry.

How this notebook grows

New topics bring more explanations, examples, diagrams, and sources to the notebook.

Original learning notes by PromptExploit, with links to TryHackMe and additional references. This independent portfolio is not an official TryHackMe certificate.

Room notes expanded · 23 September 2026

Added AI-Enhanced Attacks, Defensive AI, Securing AI, Practical, and Conclusion. The notebook now contains fifteen task notes, including the ten supplied practical questions with explained answers and three new visual guides.

First study entry · 18 September 2026

Eight task notes for The Building Blocks of AI. Two task notes for AI Security Threats. Ten notes published.

Search the entire journalesc to close