AI Red Teaming — resources
roadmap.sh: https://roadmap.sh/ai-red-teaming
Books
- Adversarial AI Attacks, Mitigations, and Defense Strategies (John Sotiropoulos) — practical playbook for attacking and hardening ML and LLM systems.
- Not with a Bug, But with a Sticker (Ram Shankar Siva Kumar & Hyrum Anderson) — accessible field guide to how ML systems fail and get attacked in the real world.
- The Developer’s Playbook for Large Language Model Security (Steve Wilson) — maps the OWASP LLM Top 10 to concrete engineering defenses.
Courses / practice
- OWASP Top 10 for LLM Applications — the canonical taxonomy of LLM vulnerabilities every red-teamer should know.
- MITRE ATLAS — adversarial threat landscape and attack-technique knowledge base for AI systems.
- Gandalf by Lakera — hands-on prompt-injection challenge to practice jailbreaking and filter bypasses.
- NIST AI Risk Management Framework — governance and risk language for framing red-team findings.