StationX
Start free

← All paths

AI Security Engineer — career path

StationX Advanced

Secure the AI a company builds and buys: LLM apps, RAG, agents, the tools they call and the pipeline and cloud they ship on. Security+ and SecAI+ for the grounding, building and red-teaming AI for the skills, then CISSP, with AAISM as the step into leading AI security.

6 courses · 62 sections · 408 lectures · 41 labs · ≈ 38 h of hands-on labs

Free to start: the free lessons and guides for this roadmap →

Try a free AI security lab: Make it talk (~5 min) → A scripted simulation in your browser. No account needed.

The path

Start here →

Step 1 — Security foundation

The security baseline every AI security role assumes.

follows the whole CompTIA Security+ (SY0-701) roadmap: 11 courses · 76 sections · 867 lectures · 148 labs · ≈ 172 h of hands-on labs

Learn

CompTIA Security+ badge

11 courses · 76 sections · 867 lectures · 148 labs · ≈ 172 h of hands-on labs

9 steps — The industry-standard entry certification for a cyber security role.

Open the full CompTIA Security+ (SY0-701) roadmap →

Step 2 — How AI applications are built

Build an LLM app, a RAG pipeline and a cloud AI service. You cannot secure what you have never built.

2 courses · 24 sections · 39 lectures · 9 labs

Learn

Practise

Free resources

Step 3 — The AI security certification

Threats, controls and governance for AI systems.

follows the whole CompTIA SecAI+ (CY0-001) roadmap: 8 courses · 110 sections · 587 lectures · 34 labs · ≈ 42 h of hands-on labs

Learn

CompTIA SecAI+ badge

8 courses · 110 sections · 587 lectures · 34 labs · ≈ 42 h of hands-on labs

6 steps — CompTIA's AI security certification: how AI works, how AI systems are attacked and secured, how to use AI in security operations, and how AI is governed.

Open the full CompTIA SecAI+ (CY0-001) roadmap →

Step 4 — Attack AI like a red teamer

How each attack on an AI system works and the open-source tools that test for it. The labs drill the web attacks AI apps inherit.

14 labs

Learn

There is no hands-on AI attack lab yet: the only prompt-injection practice is the free simulation, Make it talk. The labs below drill the web attacks AI apps inherit.

Practise

Free resources

Step 5 — Secure agents, tools and integrations

Lock down the agents, MCP servers and APIs an AI system calls, and the code that joins them.

2 courses · 20 sections · 93 lectures · 8 labs

Learn

Practise

Free resources

Step 6 — Secure the pipeline and the cloud

The courses teach Docker, Jenkins CI/CD, CloudFormation and the OWASP Top Ten. The labs secure the Azure VMs, databases and storage an AI system runs on; there is no container, infrastructure-as-code or pipeline lab yet.

2 courses · 18 sections · 276 lectures · 10 labs

Learn

Practise

Step 7 — The senior security certification

The certification AI security adverts ask for most.

follows the whole CISSP — Certified Information Systems Security Professional roadmap: 16 courses · 66 sections · 737 lectures · 71 labs · ≈ 70 h of hands-on labs

Learn

ISC2 CISSP badge

16 courses · 66 sections · 737 lectures · 71 labs · ≈ 70 h of hands-on labs

5 steps — The senior security certification, spanning eight domains. Requires five years of experience to certify fully.

Open the full CISSP — Certified Information Systems Security Professional roadmap →

Where this leads

Explore free resources

Choose your next roadmap

Advanced · after the job

AI-Driven Security Engineering

Senior roles are going to the people who can direct AI to build security solutions. When you're working in the role and ready for that step, the AI Master's Program is where it's taught — an advanced, application-only programme.

See the AI Master's Program →Application-only. A separate programme for when you're ready.