sicurity.aiTrack directory
Train the AI/ML Security Engineer.
AI security is not one isolated task. The AI/ML Security Engineer must understand applications, agents, authorized assessment, and safe automation. Each track builds one part of that connected practice.
The role
The AI/ML Security Engineer / the reason to train
AI is in the stack.
Ownership is split.
The risk does not stay inside one team. A single AI feature can touch application code, cloud services, sensitive data, model APIs, identities, and incident response.
The AI/ML Security Engineer connects those pieces, establishes controls, and stays accountable when the system behaves unexpectedly.
Know what is running
Inventory AI systems, map data paths, and find the trust boundaries.
Constrain what can act
Set identity, permissions, tool access, and limits for agents and integrations.
Test before release
Exercise applications, retrieval, models, and pipelines against realistic failure modes.
Make behavior visible
Preserve traces and evidence so teams can investigate what a system did and why.
Respond and recover
Contain unsafe behavior, protect data, restore service, and improve the controls.
This is sicurity.ai’s case for the role, not a claim that every organization has the same systems or staffing needs.
Training for the gap
Security experience
still counts.
This is a new application of familiar security work, not a reset of your career. Bring your foundation in application security, cloud, identity, detection, incident response, or governance. Build the AI-specific skills that connect it.
The goal is not to re-teach what you know. It is to help you take responsibility for systems that cross traditional team boundaries.
See how the tracks fit together ↗The role in practice
One engineer. Connected security work.
Threat-model the system, set boundaries for data and tools, test the controls, and make the result investigable. The job spans teams because the system does.
A training path cannot replace experience, organizational authority, or a security team. It can give practitioners a structured way to build and demonstrate relevant skills.
Training tracks
Current module access is listed separately from enrollment. Open a track to see each module's availability and direct Skool link when released. Planned modules are not accessible yet.
AI Platform Security Engineer
GenAI and LLM application security
Secure the models, data paths, and applications people rely on.
PROPOSED GAIPS OBJECTIVE ALIGNMENT · NOT VERIFIED COMPLETE EXAM PREPARATION
Second in rollout0 of 13 modules accessible
Agentic AI Defender
Securing agentic AI
Secure agents, their identities, tools, and authority.
PROPOSED SICURITY.AI AGENTIC DEFENSE CREDENTIAL · NOT A GIAC CERTIFICATION
First modules · Dec 14, 20260 of 13 modules accessible
Offensive AI Operator
Offensive AI
Test AI-enabled systems in authorized scope. Turn findings into defense.
SELECTED GOAA-RELATED FUNDAMENTALS · NOT COMPLETE EXAM PREPARATION
On the roadmap0 of 13 modules accessible
AI Security Automation Engineer
AI and security automation
Automate security work without automating away judgment.
PROPOSED GASAE OBJECTIVE ALIGNMENT · NOT VERIFIED COMPLETE EXAM PREPARATION
On the roadmap0 of 13 modules accessible
Further ahead: applied ML for security, Python and MCP automation, fundamentals, and DFIR tracks with GCFE and GCFA mapping. More details will accompany each course announcement.