Your AI Is an Attack Surface: What CompTIA SecAI+ Tests and How to Be Ready

Somewhere in your organization right now, a chatbot is answering customers, a copilot is reading internal documents, and an AI agent has been handed an API key. Nobody filed a change request for the attack surface that came with them. That's the gap CompTIA SecAI+ was built for, and it's why we think it's the most practical new security certification in years. Here's what it actually tests, the attacks behind every domain, and a six-week plan to be ready.

AI didn't replace your attack surface. It added one.

Security teams spent twenty years learning to defend code that does exactly what it's told. AI systems don't work that way. A model can be talked into things. Its training data can be poisoned months before anyone notices. Its outputs can leak what it was never supposed to say. And an agent with the wrong permissions can turn one bad prompt into a real transaction.

None of that shows up cleanly in a vulnerability scan, and most of it isn't covered in depth by the certifications security people already hold. CompTIA launched SecAI+ (exam CY0-001) on February 17, 2026 to close that gap, and it's aimed squarely at the people who will get the call when something goes wrong: working security practitioners.

Six attacks every SecAI+ candidate should be able to explain

CompTIA's objectives talk about adversarial attacks, model and pipeline protection, and malicious use of generative AI. Here's what that looks like on an ordinary Tuesday.

Four domains, and where the weight really is

Who should take SecAI+, and who should wait

A six-week plan to pass

1. Week 1: learn the language. Machine learning, deep learning, NLP, training versus inference, embeddings and RAG. If you can't explain how a model is trained and served, the security questions won't land.
2. Week 2: break things on purpose. Study the attack catalogs practitioners use, such as the OWASP Top 10 for LLM Applications and MITRE ATLAS. Try prompt injection against a test chatbot you control.
3. Weeks 3–4: secure the stack. This is the 40% domain. Access control for models and APIs, protecting training data and pipelines, guardrails on inputs and outputs, isolating agents and least-privilege tool access, and logging every model call.
4. Week 5: put AI to work in the SOC. Alert triage, correlation, automated playbooks and behavior analytics, and where a human must stay in the loop.
5. Week 6: governance, then drill. Read the NIST AI RMF's four functions (Govern, Map, Measure, Manage) and how GDPR applies to AI. Then do timed practice: 60 questions in 60 minutes leaves no room to wander.

SecAI+ in a security career path

• FoundationSecurity+
  The baseline most employers and DoD roles expect.
• SpecializeCySA+ or PenTest+
  Hands-on defense or offense, and both are recommended preparation for SecAI+.
• Add AISecAI+
  The AI attack surface, layered on what you already know.
• LeadCISSP or CISM, then AAISM
  When you move from defending AI systems to running the program that governs them.

Every security certification you already hold assumes the system does what it was programmed to do. SecAI+ is the first one built around systems that don't.

— Scott

Exam details are from CompTIA's published SecAI+ page as of October 12, 2026. Confirm the current version before you schedule.