AI Security Training, for professionals and security teams
A working command of AI security, not a marketing tour of it.
Practitioner-led training for the people who will actually defend your AI estate. Multi-vendor across Claude, OpenAI, AWS Bedrock and Azure AI, with hands-on labs against real tooling, MCP and agentic patterns. Built around the Claude Certified Architect (CCA-F) credential, with the $99 exam attempt sponsored.
- 74%of cyber teams are restructuring for AI (SANS 2026)
- 60%cite the skills gap as their number one challenge
- 27%had a breach tied to the skills gap
- 95%report regulatory pressure on hiring
Trained people are the spine of every maturity stage.
-
Protect
OWASP Agentic Top 10, prompt-injection defence, MCP security and guardrails.
-
Utilize
Agentic triage, hunting, vulnerability management, audit and IR, hands-on with your actual stack.
-
Govern
NIST AI RMF, ISO 42001, EU AI Act and board reporting. Speak the language.
Not just Claude: every stack your shop actually runs.
Claude (Anthropic)
Primary focus. Reasoning, code generation, security analysis, MCP, extended thinking and prompt caching.
AWS Bedrock
Enterprise deployment: VPC isolation, Guardrails, Knowledge Bases and Action Groups, the patterns regulated industries need.
Azure AI Foundry
Azure OpenAI plus Logic Apps and Defender. Where most Microsoft-shop SOCs land their first agent.
OpenAI and GPT
Broad ecosystem, legacy integrations and comparison patterns so you don't lock yourself in.
n8n and orchestration
Wire agents into existing tooling (Defender, XSIAM, Wiz, Slack, Jira) without rewriting your SOC.
Open source (Llama, Mistral)
Bonus module for air-gapped, data-residency or cost-sensitive deployments.
Individual practitioners, or a whole department.
Individual training
- For SOC analysts, threat hunters, auditors, GRC specialists and security engineers
- Self-paced modules plus live cohort labs
- Hands-on with real AI tooling, MCP and agentic patterns
- Capstone: deploy an agent against your own sample data
- Per seat, with volume discounts
Corporate cohorts
- Built for your industry, tooling and risk appetite
- 2–4 week cohorts, on-site or remote
- Three tracks on one multi-vendor backbone
- Capstone scored against the NIST AI RMF
- Invoice and L&D reimbursement letters provided
Included: a sponsored CCA-F exam attempt ($99 value)
Every qualifying engagement includes a sponsored attempt at the Claude Certified Architect, Foundations (CCA-F) exam, so your team walks away with an industry-recognised credential, not just slides. Subject to Anthropic Partner Network availability.
Three tracks. Same multi-vendor backbone.
-
SecOps
Agentic triage, hunting and IR augmentation built around your SIEM or XDR (Defender, XSIAM, Sentinel, Splunk).
Capstone: a live agent against synthetic alerts.
-
Audit & compliance
Control-evidence automation, NIST AI RMF alignment and audit narrative drafting, mapped to SOC 2, ISO and PIPEDA.
Capstone: an auditor-ready evidence packet.
-
Risk & governance
AI risk register design, policy authoring and board reporting, with NIST AI RMF, ISO 42001 and EU AI Act crosswalks.
Capstone: a board-grade AI risk report.
Three guarantees on what graduates can do.
-
Build
Graduates can design, secure and ship a working agentic workflow against your stack: MCP, guardrails, audit trail and human-in-the-loop.
-
Defend
Graduates can red-team an AI surface against the OWASP Agentic Top 10 and MITRE ATLAS, and explain the verdict.
-
Govern
Graduates can crosswalk a control set to NIST AI RMF, ISO 42001 and the EU AI Act, and brief the board.
Tell us where you are in your AI journey.
Individuals, teams or whole departments use the same form. We match you to the right tier or design something custom.
Do I need to be a developer?
No. The curriculum is built for security practitioners: SOC analysts, threat hunters, auditors and GRC specialists. You will write and read some configuration and light code, but the labs are designed around security problems, not software engineering.
Is this just a Claude course?
No. The backbone is deliberately multi-vendor: Claude, OpenAI, AWS Bedrock, Azure AI Foundry, n8n and open-source models. Same patterns, different SDKs, so you learn when to reach for which.
What is the CCA-F exam attempt?
Claude Certified Architect, Foundations is an Anthropic credential. Every qualifying seat includes one sponsored attempt, a $99 value, and the labs are aligned to the exam blueprint.
Can you run it on-site?
Yes. Corporate cohorts run on-site or remote over 2–4 weeks, scoped to your team. We provide invoices and L&D reimbursement letters.