Red-team report: attacking the support assistant before a customer does

A security review asks one question about an AI assistant: what happens when someone tries to break it? This project answers with numbers from real attacks, before and after a fix.

The brief

The support assistant is about to go live on a public website. The customer's security team will not approve it without a report showing it was attacked, what got through, and what was done about it.

What you will build

  • A garak scan of the assistant for prompt injection, jailbreaks and data leakage
  • Two PyRIT attacks written for this assistant, such as extracting another customer's order or getting a refund promised
  • A fix for the worst finding, in guardrails or in the tools
  • A second run showing the attack success rate after the fix
  • A two-page report for a security reviewer

Requirements

RequirementDone when
Scangarak runs against the assistant itself, not a bare model, and its report is kept
Targeted attacksAt least two PyRIT attacks specific to this assistant's tools and data
ScoringEach attack's success is judged by a scorer, and the scorer was checked on a few examples by hand
FixOne finding is fixed and the same attacks are run again
ReportStates attack success rates before and after, what is still open, and the risk accepted

What to learn it from

Milestones

  1. Wrap the assistant so garak and PyRIT can call it
  2. Run the garak scan and record the baseline
  3. Write and run the two targeted attacks
  4. Pick the worst finding and fix it
  5. Run everything again with the same settings
  6. Write the report
Only attack what you own
Run these tools against systems you built or have written permission to test. Scanning someone else's assistant is an attack, not a test.

Stretch goals

  • Fail the build when the scan's attack success rate rises
  • Add an indirect prompt injection hidden in a document the assistant retrieves
  • Measure how much the fix slowed down normal answers

Before you call it done

Worth remembering
  • The same attacks were run before and after the fix
  • Every number in the report came from a saved run
  • The report says what is still open, not only what was fixed
Back toAll frameworks

Every expert started right here.