AI Engineer Bootcamp: The OWASP Top 10 for LLM Applications: From Risks to Mitigations

Cloud Security Engineer Bootcamp: The OWASP Top 10 for LLM Applications: From Risks to Mitigations

Learn how to identify, test, and mitigate real security risks in modern AI systems through a hands-on course focused on attack paths, defensive controls, and measurable outcomes.

About this course

AI systems introduce new security challenges that traditional application security does not fully address. Prompt injection, data exfiltration, tool misuse, unsafe retrieval, hallucination risk, and agent abuse all require new defensive patterns and better ways to evaluate whether protections actually work.

This course is built around a simple model: attack, defend, and measure. You will explore how AI systems can fail, implement practical mitigations, and assess those mitigations using security-focused evaluation techniques. The course structure follows widely adopted cybersecurity training practices that emphasize clear outcomes, hands-on activities, and straightforward navigation in online learning environments.

What you’ll learn

By the end of this course, you will be able to:

Recognize major security risks in LLM applications and agent-based systems.
Apply defensive controls to prompts, tools, retrieval pipelines, and outputs.
Improve resilience against injection, exfiltration, poisoning, evasion, and misuse.
Investigate incidents using audit trails, traces, and operational signals.
Measure security readiness using practical evaluations, scorecards, and release gates.

Who this course is for

This course is intended for:

Security engineers and architects
Software developers
DevSecOps practitioners
Technical leaders responsible for AI and application securitygenai.owasp+1

It is especially well suited for people building or reviewing LLM applications, copilots, and autonomous or semi-autonomous agent workflows.

Course approach

The course is hands-on and lab-driven, with each module centered on a specific security problem and a practical way to test or mitigate it. You will work through real attack paths, implement defensive controls, and use security evaluations to see which protections actually hold up under pressure.

Richard Augenti, AI Security Architect

Richard Augenti is an AI Security Architect with decades of experience across security, information technology, DevSecOps, and AI security. His work brings together technical depth and practical security engineering, with a focus on building AI systems that are secure, resilient, and ready for real-world deployment.