Applied Data Science and AI for Cybersecurity at Black Hat USA 2026

The Black Hat USA 2026 sessions are finished. This course ran August 1 to 4, 2026 in Las Vegas. It is available now as private on-site or virtual training. The four-day AI Cyber Bootcamp, which covers this material and more, runs publicly at Black Hat India 2026, October 27 to 30.

Data science, applied to your security data

Most security teams sit on more data than they can use: PCAP, EDR telemetry, authentication logs, phishing samples, alert queues. Applied Data Science and AI for Cybersecurity teaches you to turn that data into detection and analysis using the same tools working data scientists use, on the security problems you actually face.

This is GTK Cyber’s flagship course. At Black Hat USA 2026 in Las Vegas it ran as two identical 2-day sessions, August 1 to 2 and August 3 to 4, so students could fit it around the rest of the conference week. Both sessions covered the same material.

To run it for your own team, on your dates and against your own data, get in touch.

Who it is for

The course targets intermediate security practitioners who want practical data science skills: SOC analysts moving toward detection engineering, threat hunters who want statistical and machine learning methods, and security engineers preparing for AI work in the SOC. If you know your security domain and can read Python, this course starts where you are.

What you work through

Across four days (32 hours, half instruction and half labs in the AI Training Dojo):

• Loading and cleaning messy security data with Pandas
• Feature engineering from logs, flows, and telemetry
• Data visualization for security investigation
• Supervised learning for phishing, malware, and intrusion classification
• Unsupervised learning and anomaly detection for hunting
• Model evaluation, optimization, and where large language models fit
• A first look at adversarial machine learning: how models get evaded and poisoned

Every concept is followed by a lab on real security data. You keep the notebooks.

Read first, then decide

If you want a sense of the work before you commit a seat, these go deeper on techniques from the course:

• Data Science for Incident Responders
• Building a Threat Hunting Pipeline with Python and Jupyter
• Reducing False Positives with Machine Learning

Bringing a team?

Group rates are available for three or more. Contact [email protected] before registering. If a public course does not fit, GTK Cyber also runs this material as private on-site training.