CPP-03
Secure AI Apps & Agents
A no-jargon look at what actually goes wrong with AI apps and agents — prompt injection, agent permission risk, real examples — and what "secure by design" looks like.
What you'll learn
- Recognise the real ways AI apps and agents get compromised or misused
- Understand agent identity and permission risk in plain terms
- Know the questions to ask before an AI agent gets real tool access
- Know where to go next — the full Cloud & AI Security Engineering programme
Full curriculum — 1 day
- S1 How AI apps actually get attacked — prompt injection & jailbreaks, in plain terms
- S2 Agent identity & permission risk — what changes once AI can take actions
- S3 A secure-by-design checklist for a real AI app or agent
- S4 Where next — from workshop to Cloud & AI Security Engineering
How it's taught
Every session is live and taught personally by Muyiwa Ogunlade, a Microsoft Certified Trainer. Cohorts are capped so there is room for discussion, code review and direct feedback. Recordings are provided for revision — they are a backup, not the delivery method.
Teaching runs Monday to Thursday, 18:30–21:00, and Saturday mornings 09:30–13:00 (UK time), so the programme fits around a full-time job. West Africa Time runs 0–1 hour ahead, which makes the evening sessions workable from Nigeria as well as the UK. See all the ways to train.
Book your place
Next step
Ready to start CPP-03?
Propose a start date at least 10 days out. If the trainer is available, we open it as a public live-online cohort other learners can join.