Get certified. Hands-on AI security.
The AISEC Practitioner course equips IT security professionals to defend AI systems in production. Six domains. Six labs. NIST AI Risk Management Framework aligned. Two certification levels.
See the curriculum ↓Why take it
Three reasons this course is different
01
Hands-on, not theoretical
Six labs across prompt injection, backdoor attacks, image manipulation, LLM Guard, vulnerability scanning, and agentic-AI red-teaming. You will execute the attacks and apply the controls, not just read about them.
02
NIST AI RMF aligned
Risk management curriculum maps to the NIST AI Risk Management Framework, the standard regulators are converging on. Walk out able to apply the framework to live AI deployments — not just describe it.
03
Built on real client engagements
The threats, controls, and labs are drawn from Outpace's Lastmile work defending live AI deployments inside Australian enterprises. You will learn the failure modes that actually happen in production — not the ones that look good on a slide.
The curriculum
Six domains. Six labs. Working knowledge.
Each domain combines a knowledge component (concept, risk, control) with a skills component (a hands-on lab where you apply it). The structure mirrors the way a real AI security review is run — diagnose, control, verify.
AI Technology
- History and basic concepts of AI; model architecture.
- Tokenisation and embedding for an AI model.
- How data is used to train and enhance AI models.
- Run AI models locally, from marketplaces, and via online services.
- Run AI models programmatically.
AI Risk Management
- Threat identification across the AI stack.
- Prompt injection techniques to extract sensitive data.
- Inject backdoors into AI models and execute shells.
- Manipulate images to defeat classification models.
- Apply the NIST AI Risk Management Framework.
AI Governance
- Governance approach — frameworks, accountability, decision rights.
- Conceptual AI architecture and information architecture.
- Review and author an AI policy document.
AI Controls
- Key controls that protect AI systems in production.
- Apply guardrails — common design patterns and pitfalls.
- Red-team AI with model scanners.
- Logging and monitoring for AI systems.
Securing AI Agents
- Principles of agentic AI.
- Agentic AI mesh architecture.
- Develop an AI agent using the smolagent framework.
- Assess the security implications of AI agents.
Hands-on AI Labs
- Tokenisation, embedding, ML characterisation.
- Prompt and thought injections; the Gandalf Challenge.
- LLM Guard, vulnerability scanning, shell-code injection.
- Application monitoring and incident response for AI.
Who it's for
IT professionals defending AI in production
Designed for security and engineering roles where AI is now part of the attack surface. Risk and security leaders extending into AI get the most leverage; SecOps, MLOps, and DevOps engineers pick up the day-two skills fastest.
Typical attendee roles
- AI security professionals
- SecOps, CloudOps, NetOps, MLOps, DevOps Engineers
- IT Security Professionals, Practitioners, Architects
- IT Support Staff and Managed Service Providers
- Product Managers, Software Engineers, Testers
- Data Governance Leaders, Data Architects
- SOC Analysts, Penetration Testers, Security Engineers
Upcoming sessions
Pick a session that fits your schedule
Sessions run live with the instructor on Zoom (or in person in Melbourne). Same syllabus, same exam, same certification.
12–14 August
Virtual · 9am–5pm AEDT
Face-to-face: $3,250 ex GST · Melbourne
9–11 November
Virtual · 9am–5pm AEDT
Face-to-face: $3,250 ex GST · Melbourne
Group bookings
Want to bring your team?
We offer tailored rates for five or more attendees from the same organisation. Just reach out and we'll put something together.
Certification
Two levels. Same course.
The exam is a two-part online assessment taken during the course. Pass the multiple-choice for Level 1; complete the labs for Level 2.
AISEC — Practitioner
Pass the multiple-choice exam: 30 questions, 1 hour, 75% pass mark. Demonstrates working knowledge of AI risk, governance, and controls.
AISEC+ — Practitioner with Labs
Pass the multiple-choice plus at least three of four hands-on lab assessments: secure AI design, threat assessment, guardrails, agentic-AI security. The stronger signal for hiring managers.
One free resit available for either component if needed.
FAQ
Common questions
FREQUENTLY ASKED
Frequently Asked Questions
Outpace AI authors the syllabus, exam, and lab content. Live delivery and enrolment are handled by ALC Training, Outpace's certified training partner across ANZ. You attend an ALC-run session and receive an Outpace-issued certification.
Both. Virtual sessions run 9am–5pm AEDT live with the instructor on Zoom and the exam invigilated live online. Face-to-face sessions run in Melbourne with the exam paper-based and invigilated in the same room. Pick whichever fits your schedule — same syllabus, same certification.
There are two levels. Level 1 (AISEC) — pass the 30-question multiple-choice exam (75% pass mark, 1 hour). Level 2 (AISEC+) — pass the multiple-choice plus at least three of four hands-on assessments simulating secure AI design and threat assessment. Both levels demonstrate working knowledge of AI risk, governance, controls, and agentic-AI security in production environments.
The course is designed for IT professionals with a working background. Existing AI systems knowledge and basic Python familiarity are an advantage but not prerequisites. Risk and security leaders extending into the AI domain are the most common attendees, but the curriculum is accessible to SecOps, MLOps, DevOps engineers and IT architects too.
One free resit is available for either component (multiple-choice or hands-on). Course material updates regularly to track AI developments, so resits should be taken as quickly as practicable.
Yes — five or more attendees from the same organisation qualify for a custom group rate. Get in touch for a quote and to discuss scheduling a private session for your team.
AI Fluency is a free self-paced course on governed AI use for general workforce — the 4 Ds framework, ~3 hours, no specialist prerequisites. AI Security (AISEC) is a paid, instructor-led, certification-bearing 3-day course for IT security professionals: model architecture, prompt injection, red-teaming, NIST AI Risk Management Framework, agentic-AI security. Different audiences, different depths.
Live instructor delivery across 3 days, digital course workbook, supplementary materials, lab access, certification exam voucher (both Level 1 and Level 2 attempts), and the resit option if needed.
Ready to certify?
Book the next AISEC session
Three days. Two certification levels. Practical AI security skills you can use the moment you walk out of the course.