BBitMonkey AI
Home / Case studies / T11
T11The Fourteen

Rangoli Retail

Red-teaming an AI system before launch

Sign in to open this case study.

Sign in

The case

An online retailer's AI customer-service assistant is ready to launch. The team has demonstrated it answering thirty questions perfectly. A red-teamer is given two weeks, a staging environment and a written set of rules.

What is in it

  • 12 teaching steps, one animation each
  • 29 labs, checked live against the case data
  • 60 step questions, every option explained
  • 22 final paper questions, marked on submission
  • 12 Claude prompts, optional

The steps

  1. 1The launch that nearly happened · BEFORE YOU TEST
  2. 2Rules of engagement · BEFORE YOU TEST
  3. 3The attack taxonomy · BEFORE YOU TEST
  4. 4Grounding · THE PROBES
  5. 5Prompt injection · THE PROBES
  6. 6Personal data · THE PROBES
  7. 7Tool actions · THE PROBES
  8. 8Severity and likelihood · RANK AND FIX
  9. 9What the blue team built · RANK AND FIX
  10. 10Retest, and the cost of the fix · LAUNCH
  11. 11The launch gate · LAUNCH
  12. 12Monitoring and the next cycle · LAUNCH