Special price: USD 69 — available only until 6 November 2026.

Understand the reasoning behind AI-powered security decisions

Explainable AI in Cybersecurity introduces the concepts and practical workflows that help security professionals move from opaque alerts to informed, human-led decisions. Learn to evaluate AI-generated evidence, understand model explanations, and document the reasoning behind your response.

Instructor: Cordny Nederkoorn
Format: Online course with five modules, knowledge checks, practical exercises and a final multiple-choice exam.
Estimated module workload: 60–80 minutes, excluding the introduction and final exam.
Start date: 30 October 2026.

What you will learn

• Recognize the limits of black-box models, confidence scores and automated alerts.
• Distinguish explainability from interpretability, global from local explanations, and model-agnostic from model-specific approaches.
• Understand false positives, false negatives, information bias and automation bias.
• Apply the five-step workflow: Detect, Explain, Evaluate, Decide, Document.
• Connect explainability with accountability, documentation, auditability and human oversight.
• Practice counterfactual explanations in a cybersecurity context.

Who should attend?

SOC analysts, incident responders, security engineers and architects, compliance managers, and CISOs interested in understanding and evaluating AI-supported security decisions.

Prerequisites and equipment

A fundamental understanding of SOC detection and response, plus basic awareness of machine-learning models, training data and risk scoring. Advanced data-science programming is not required. You will need a laptop or desktop computer; an additional monitor is optional.

Course syllabus

Module 0: Before the course

Meet the instructor and review the course outline.

Module 1: Trust and Transparency in Cyber-AI

Explore why AI is increasingly used in security operations, why traditional models can be difficult to understand, and why a high confidence score does not necessarily mean a trustworthy decision. Examine the human–machine gap, false positives, false negatives and automation bias.

Workload: 10 minutes.

Module 2: What is XAI?

Explainability versus interpretability; global versus local explanations; feature importance; model-agnostic versus model-specific approaches; and where XAI helps—and where it does not.

Exercise: Knowledge check on explainability.
Workload: 10–15 minutes.

Module 3: XAI — Governance and Regulation

Examine explainability in accountability and risk management, the EU AI Act and NIS2 from a cybersecurity perspective, documentation, auditability and human oversight. Discuss why “the AI decided it” is not an adequate operational explanation.

Exercise: Knowledge check on the regulations covered.
Workload: 15–20 minutes.

Module 4: From AI Decision to Human Decision

Use a practical five-step XAI cybersecurity framework to detect, explain, evaluate, decide and document. Combine machine-generated evidence with human context and judgment.

Exercise: Apply the framework to a realistic SOC alert.
Workload: 15–20 minutes.

Module 5: Counterfactual Explanations

Learn what counterfactual explanations are and practice using this XAI technique.

Exercise: Practice counterfactual explanations.
Workload: 10–15 minutes.

Final exam

A multiple-choice exam covering all modules.

About the instructor

Cordny Nederkoorn — Software Quality, Explainable AI & Cybersecurity

Cordny helps make complex technology understandable, trustworthy and fit for purpose. With over 20 years of experience in software testing and quality engineering, he combines quality engineering with technical documentation and a growing focus on Explainable AI.

Through TestingSaaS, he explores how techniques such as SHAP and LIME can help organizations understand AI-driven decisions, investigate potential bias and strengthen human oversight. His cybersecurity interests lie at the intersection of XAI, responsible AI and digital trust.

His mission is to bridge the gap between complex technology and practical understanding, helping teams turn technical knowledge into better decisions, more transparent systems and higher-quality software.