About Course

Beidat Academy · Graduate Certificate

AI in Cybersecurity

From machine learning basics to adversarial attacks, LLM security, and AI governance — everything you need to work at the intersection of AI and cybersecurity.

📚 65+ Lessons🧪 13 Hands-on Labs✅ 13 Quizzes🆓 Free Lab Access⏱️ Self-Paced

What You Will Learn

🤖 AI for Security
Use machine learning to detect intrusions, classify malware, and spot anomalies in logs — the same way real security teams do it.

🛡️ Security of AI
Understand how AI models can be attacked, fooled, or poisoned — and how to defend them.

🗡️ Adversarial Machine Learning
Learn how attackers fool AI models with evasion and poisoning attacks — and the defenses that stop them.

🤖 LLM Security
Master prompt injection, jailbreaks, RAG security, and agentic AI risks — the hottest area in AI security right now.

📋 AI Governance
Navigate NIST AI RMF, EU AI Act, bias, fairness, and explainability requirements for enterprise AI.

🔧 Securing AI Systems
MLOps security, model cards, supply chain risks, and Zero Trust architecture for AI infrastructure.

Course Structure

Section	Topics
Weeks 0–2	Python & ML Foundations — Getting started in Colab, ML basics, confusion matrix, and metrics
Weeks 3–5	AI for Security — Intrusion detection, malware classification, anomaly detection & UEBA
Midterm	25-question exam covering Weeks 1–5
Weeks 6–8	Adversarial ML & LLM Security — Attacks, defenses, prompt injection, RAG, and agentic AI
Weeks 9–10	AI Governance & Deployment Security — NIST RMF, EU AI Act, MLOps, Zero Trust
Final	35-question final exam + Capstone project

🎓

No prior AI experience needed.

This course starts from scratch. If you know basic Python and have curiosity about security, you are ready to begin. Labs run in Google Colab — free, no installation, just sign in with your Gmail.

Course Content

Week 0 — Getting Ready
A primer for mixed-background students. Python basics, confusion matrix, and Colab. No grade weight.

Chapter: What Is Machine Learning, Really?
Watch: ML and Colab Explained
Watch: Your First Colab Notebook
Week 0 Practice Check
Slides: Course Orientation & Tools Setup

Week 1 — The AI–Security Landscape
The two directions: AI for security and security of AI. MITRE ATLAS v2026.09. Why learning systems change the threat model.

Week 2 — Machine Learning Foundations for Security Data
Features, labels, train/test splits. Why accuracy is a lie on imbalanced data. Precision, recall, F1, base rate fallacy.

Week 3 — Machine Learning for Intrusion Detection
NSL-KDD, feature groups, the three detection paradigms, and the five reasons published results rarely survive deployment.

Week 4 — Malware Detection and Classification
Static vs dynamic analysis. PE headers as features. Packing and obfuscation. Concept drift in malware classifiers.

Week 5 — Anomaly Detection, Logs and UEBA
When you have no labels. Isolation Forest. Log parsing and templating. Why anomalous ≠ malicious. UEBA risk scoring.

Midterm Assessment — Weeks 1–5
25-question exam covering Weeks 1–5. Two attempts, 45-minute limit, 70% pass. Counts 15% of final grade.

Week 6 — Adversarial Machine Learning
Evasion (FGSM), poisoning (backdoors), model extraction, membership inference. NIST AI 100-2 taxonomy. Defences that work vs defences that look like they work.

Week 7 — LLM Security I: Prompt Injection and the OWASP LLM Top 10
LLM01–LLM10 (OWASP 2026). Prompt injection anatomy. How RAG and agents change the attack surface.

Week 8 — LLM Security II: RAG, Agents and Red Teaming
RAG poisoning, agentic AI risks, real-world LLM application pen testing. PortSwigger Web Security Academy LLM labs.

Week 9 — AI Governance and Compliance
EU AI Act, NIST AI RMF, ISO 42001. Risk categories, prohibited practices, transparency obligations, conformity assessment. How governance changes what security teams must deliver.

Week 10 — Securing AI Systems End-to-End
MLOps pipeline security, supply chain, model cards, SBOMs, incident response for ML systems. Capstone briefing.

Final Exam and Capstone
35-question comprehensive final exam plus capstone project submission. Counts 25% of course grade. All 10 weeks covered.

Student Ratings & Reviews

No Review Yet

AI in Cybersecurity

About Course

AI in Cybersecurity

What You Will Learn

Course Structure

Course Content

Week 0 — Getting Ready A primer for mixed-background students. Python basics, confusion matrix, and Colab. No grade weight.

Chapter: What Is Machine Learning, Really?

Watch: ML and Colab Explained

Watch: Your First Colab Notebook

Week 0 Practice Check

Slides: Course Orientation & Tools Setup

Week 1 — The AI–Security Landscape The two directions: AI for security and security of AI. MITRE ATLAS v2026.09. Why learning systems change the threat model.

Chapter: The AI-Security Landscape

Watch: AI Attacks and the 2026 Threat Index

Watch: Exploring the AI Cybersecurity Threat Landscape

Week 1 Check

Slides: Introduction to AI and Machine Learning

Week 1: Threat Landscape Research Report

Week 2 — Machine Learning Foundations for Security Data Features, labels, train/test splits. Why accuracy is a lie on imbalanced data. Precision, recall, F1, base rate fallacy.

Watch: Confusion Matrix, Imbalanced Data, and Overfitting

Watch: Building and Measuring an Intrusion Detection Model

Week 2 Check

Slides: Machine Learning Fundamentals for Security

Week 2: ML Metrics Analysis

Week 3 — Machine Learning for Intrusion Detection NSL-KDD, feature groups, the three detection paradigms, and the five reasons published results rarely survive deployment.

Chapter: Network Intrusion Detection

Watch: Detection Architecture and SIEM

Watch: Build a Log Analyzer with Python

Watch: Explore a Free SIEM Tool — Splunk

Week 3 Check

Slides: AI-Powered Threat Detection

Week 3: IDS Design Proposal

Week 4 — Malware Detection and Classification Static vs dynamic analysis. PE headers as features. Packing and obfuscation. Concept drift in malware classifiers.

Chapter: Malware Classification with AI

Watch: Malware and Machine Learning

Watch: Detecting Malware Behavior with ML

Week 4 Check

Slides: Malware Detection with AI

Week 4: Malware Analysis Write-Up

Week 5 — Anomaly Detection, Logs and UEBA When you have no labels. Isolation Forest. Log parsing and templating. Why anomalous ≠ malicious. UEBA risk scoring.

Chapter: Anomaly Detection

Watch: User Behavior Analytics and Anomaly Detection

Watch: Anomaly Detection with Isolation Forest

Week 5 Check

Slides: User Behavior Analytics

Week 5: Anomaly Detection Reflection

Midterm Assessment — Weeks 1–5 25-question exam covering Weeks 1–5. Two attempts, 45-minute limit, 70% pass. Counts 15% of final grade.

Midterm: Instructions and Review Guide

Midterm Examination

Slides: Midterm Review

Week 6 — Adversarial Machine Learning Evasion (FGSM), poisoning (backdoors), model extraction, membership inference. NIST AI 100-2 taxonomy. Defences that work vs defences that look like they work.

Chapter: Adversarial Machine Learning

Watch: Adversarial Attacks and Defences

Watch: Create and Detect Adversarial Examples

Week 6 Check

Slides: Adversarial Machine Learning

Week 6: Adversarial Attack Journal

Week 7 — LLM Security I: Prompt Injection and the OWASP LLM Top 10 LLM01–LLM10 (OWASP 2026). Prompt injection anatomy. How RAG and agents change the attack surface.

Chapter: LLM Security and Prompt Injection

Watch: Prompt Injection and LLM Security

Watch: Testing Prompt Injection Attacks

Watch: Building LLM Safety Filters

Week 7 Check

Slides: LLM Security and Prompt Injection

Week 7: Prompt Injection Report

Week 8 — LLM Security II: RAG, Agents and Red Teaming RAG poisoning, agentic AI risks, real-world LLM application pen testing. PortSwigger Web Security Academy LLM labs.

Chapter: Red Teaming and Offensive AI

Watch: Agentic AI Security

Lab: Agentic AI Security — Finding Weaknesses

Lab: Building AI Agent Safety Filters

Week 8 Check

Slides: Agentic AI and Autonomous System Security

Week 8: Agentic AI Risk Assessment

Week 9 — AI Governance and Compliance EU AI Act, NIST AI RMF, ISO 42001. Risk categories, prohibited practices, transparency obligations, conformity assessment. How governance changes what security teams must deliver.

Chapter: AI Governance and Ethics

Watch: EU AI Act and AI Governance

Lab: EU AI Act Risk Assessment

Week 9 Check

Slides: AI Ethics, Bias, and Regulatory Compliance

Week 0 — Getting Ready
A primer for mixed-background students. Python basics, confusion matrix, and Colab. No grade weight.

Week 1 — The AI–Security Landscape
The two directions: AI for security and security of AI. MITRE ATLAS v2026.09. Why learning systems change the threat model.

Week 2 — Machine Learning Foundations for Security Data
Features, labels, train/test splits. Why accuracy is a lie on imbalanced data. Precision, recall, F1, base rate fallacy.

Week 3 — Machine Learning for Intrusion Detection
NSL-KDD, feature groups, the three detection paradigms, and the five reasons published results rarely survive deployment.

Week 4 — Malware Detection and Classification
Static vs dynamic analysis. PE headers as features. Packing and obfuscation. Concept drift in malware classifiers.

Week 5 — Anomaly Detection, Logs and UEBA
When you have no labels. Isolation Forest. Log parsing and templating. Why anomalous ≠ malicious. UEBA risk scoring.

Midterm Assessment — Weeks 1–5
25-question exam covering Weeks 1–5. Two attempts, 45-minute limit, 70% pass. Counts 15% of final grade.

Week 6 — Adversarial Machine Learning
Evasion (FGSM), poisoning (backdoors), model extraction, membership inference. NIST AI 100-2 taxonomy. Defences that work vs defences that look like they work.

Week 7 — LLM Security I: Prompt Injection and the OWASP LLM Top 10
LLM01–LLM10 (OWASP 2026). Prompt injection anatomy. How RAG and agents change the attack surface.

Week 8 — LLM Security II: RAG, Agents and Red Teaming
RAG poisoning, agentic AI risks, real-world LLM application pen testing. PortSwigger Web Security Academy LLM labs.

Week 9 — AI Governance and Compliance
EU AI Act, NIST AI RMF, ISO 42001. Risk categories, prohibited practices, transparency obligations, conformity assessment. How governance changes what security teams must deliver.

Week 10 — Securing AI Systems End-to-End
MLOps pipeline security, supply chain, model cards, SBOMs, incident response for ML systems. Capstone briefing.

Final Exam and Capstone
35-question comprehensive final exam plus capstone project submission. Counts 25% of course grade. All 10 weeks covered.