Practical AI security training

Train for the work AI is changing.

Autonomous AI can already find vulnerabilities and run agentic tooling. What separates a security professional now is judgment — knowing what to trust, what to verify, and when the AI is wrong. Our hands-on courses teach and assess exactly that skill.

Talk to us about training See our assessments

Learners perform real cybersecurity work with real tools, including AI. We teach and assess both execution and the judgment behind it.

Free course

AI Security Foundations

Learn how AI agents reason, use tools, and create new security risks. Set up a safe lab, investigate a realistic agent failure, and apply your first containment control.

No AI-security experience required.

Start the free course

Continue into four specialist paths

Course 1

AI-Assisted Vulnerability Discovery, Remediation, and Validation

Run an agentic application-security pipeline against real source code — from attack-surface mapping and threat modeling to exploitability verification, remediation, and adversarial validation. Challenge AI-generated findings, reject false positives, review proposed patches, and prove the original attack path is closed.

Built for

Application security engineers, penetration testers, DevSecOps engineers, and vulnerability-management analysts.

What you’ll do

  • Map the application attack surface
  • Build a threat-informed testing plan
  • Direct specialist vulnerability agents
  • Verify exploitability and reject false positives
  • Review AI-generated remediation
  • Adversarially validate the fix

Capstone: Find, reproduce, remediate, and close a real attack path.

Start course
Course 2

Securing Agentic AI — Containment, Detection and Incident Response

Threat-model, contain, monitor, investigate, and safely restore environments that run autonomous or semi-autonomous AI agents.

Built for

Cloud and platform security engineers, SOC and incident-response analysts, detection engineers, and AI infrastructure teams.

What you’ll do

  • Reconstruct an agentic security incident
  • Threat-model an AI-agent environment
  • Build containment controls
  • Detect autonomous boundary violations
  • Contain and investigate
  • Validate containment

Capstone: The Agent Escaped.

Start course
Course 3

AI-Driven Penetration Testing of Your Own Network

Plan, authorize, and run an AI-agent-led penetration test against your own network — direct the recon and exploitation, keep the agent inside scope, and turn what it finds into a defensible report before an adversary’s agent finds it first.

Built for

Penetration testers, red teamers, internal security teams, and IT admins responsible for their own network.

What you’ll do

  • Scope and authorize an agentic pentest
  • Direct AI-driven reconnaissance
  • Supervise autonomous exploitation
  • Keep the agent inside the rules of engagement
  • Validate and reproduce findings
  • Report and remediate

Capstone: Your network, tested before theirs is.

Start course
Course 4

Detecting and Defending Against Coordinated AI Agent Swarms

Run tabletop exercises and purple-team operations against realistic, self-organizing agent swarms in your own infrastructure — then prove whether your SOC can see, understand, and contain their coordinated behavior.

Built for

SOC analysts, detection engineers, incident responders, threat hunters, purple teams, and security leaders responsible for operational readiness.

What you’ll do

  • Threat-model coordinated agent swarms
  • Design a swarm incident tabletop
  • Instrument cross-agent activity
  • Detect coordination across identities and hosts
  • Purple-team containment playbooks
  • Measure and close SOC visibility gaps

Capstone: The Swarm Is Already Inside.

Start course

From guided training to independent proof

  1. 1

    Guided modules

    Prepared findings and agent artifacts, real terminal labs, and a concept explained before every task.

  2. 2

    Advanced modules & capstones

    Complete live runs against small synthetic applications and isolated agent environments.

  3. 3

    Independent practical assessment

    The same skills, no guidance — a deterministic, evidence-backed outcome is required.

  4. 4

    Employer-facing skills report

    Verified proof of execution and judgment, ready to share with a hiring team.

Bring this to your team

Every module pairs a concise concept with a guided terminal task, real security and AI tooling, and deterministic checkpoints — so you know execution happened and judgment was tested.

Talk to us