Rules of Engagement for AI Red Teamings
Rules of engagement are the written agreement that defines what a red team may test, how, when and with what data. They set scope, authorisation, off-limits systems, reporting channels and ethical limits, protecting both the tester and the organisation from harm or legal trouble.
Stage: Foundations · Part of the AI Red Teaming roadmap · Last reviewed 19 November 2026
What to learn in Rules of Engagement
- Write a scope statement listing in-scope models, endpoints and data
- Obtain written authorisation before running any test
- Define stop conditions for harmful or sensitive output
- Handle discovered personal data or harmful content responsibly
- Agree on how and when to disclose findings
Rules of Engagement interview questions
Why is written authorisation essential?
Testing without permission can break laws and terms of service, even with good intent. Written authorisation names the scope, the dates and the people responsible, so everyone knows the activity is legitimate and what is allowed.
What should you do if testing exposes real personal data?
Stop probing that path, avoid copying or storing the data, and report it through the agreed channel straight away. Record only the minimum evidence needed to show the issue, and follow the data-handling rules in your agreement.
Where Rules of Engagement fits in the AI Red Teaming roadmap
Rules of Engagement belongs to the Foundations stage (intermediate level), which takes about 15 hours. Threat landscape.
Other units in this stage
Back to the full AI Red Teaming roadmap