Track 5
Security engineering for AI systems
The security basics every AI role is now expected to know: threat models, injection, data protection, identity and supply chain.
Securing AI systems
Threat modelling, prompt injection and tool permissions, data protection, identity and supply chain, and a review checklist.
- 01Threat modelling an LLM applicationDraw the data flows, mark every place untrusted text meets a model or a tool, and walk the OWASP list against each.
- 02Prompt injection and tool permissionsYou cannot reliably filter injection out of text, so design so that a fooled model still cannot do damage.
- 03Data protection, tenancy and retentionKnow what data enters prompts, logs, indexes and caches, and keep each tenant's data in its own lane everywhere it goes.
- 04Identity, secrets and supply chain for agentsAgents are workloads with identities: give them short-lived, scoped credentials and treat models, tools and MCP servers as dependencies.
- 05An AI feature security reviewA checklist you can run in a design review or offer in an interview, with the reasoning behind each line.