Skip to main content

Red-team an AI system

Plan and run an authorised red-team exercise on an AI app: scope, rules of engagement, manual and automated attacks with PyRIT, garak and promptfoo, attack success rate, findings and retests.

Pro 2 · Production engineerabout 4 hours5 sections

Quick answer

AI red teaming is authorised testing where people and tools play the attacker against your own AI system, measure how often attacks succeed, report findings and retest after fixes.

Why it matters

Every serious AI launch now needs evidence that someone tried to break it first. Red teaming is a core skill for AI security roles, and a written red-team report is strong proof of that skill in an interview.

Sign up free to read the whole lesson

The rest of this lesson: 5 parts, video lessons, the best course, practice and a quick check. It is free, takes a few seconds and needs no card.

Accounts are for learners aged 18 and over. Under 18? The quick answer and the first part of every lesson stay open: browse lessons by topic.

On the roadmap: AI Security Engineer

SecurityPro 2: Production engineerVideo lessonsRed teamingJailbreakPrompt injectionPyRITgarakData exfiltration

Track

Today's AI turns are used

Lessons, roadmaps, videos, practice cards and the code playground stay free with your account, with no limit.