Skip to content
AI Security

Interview question · AI Red Teaming, Evaluation & Response

Two teams each claim their prompt injection defence is better. How do you compare them fairly?

Compare actual configurations with paired evidence, separate adaptive budgets and useful-task checks. A lower rate or significant test does not settle the release decision.

hardNewUpdated Nov 2026

Compare actual configurations with paired evidence, separate adaptive budgets and useful-task checks. A lower rate or significant test does not settle the release decision.

Make your first attempt

What would you say before reading the answer?

  1. 01 State the assumptions your answer depends on.
  2. 02 Explain the mechanism and choose a control.
  3. 03 Name a condition that would change your decision.

The worked answer includes an approach, scoring criteria and follow-up questions. Compare the reasoning with your own.

10 answers per topic instead of 5, plus saved progress and bookmarks · costs nothingor unlock all 529 remaining answers · $25

Every worked answer in all 17 topics, every lesson in every course, the full concept curriculum, the complete practice-test bank, and the downloadable security field guides.

6 months · One payment · No auto-renewal

What if · the follow-ups in this answer

The answer works through 4 follow-ups like an interviewer would

Two document-assistant defenses produce a paired fixed-case table and separate adaptive campaigns. Candidate B has a lower observed fixed-case release rate but still has confirmed prohibited releases.

  1. What if only each candidate’s total failures were saved?
  2. What if each adaptive campaign generated different messages?
  3. What if B’s improvement is statistically significant but its releases violate a mandatory boundary?
  4. What would support choosing a candidate?

Each rung has a nudge, a model answer, the principle that carries over and the weak answer to avoid. Premium opens every answer and ladder.

UP NEXT ON YOUR JOURNEY
Next in this trackHow do you preserve evidence in an AI incident when the logs contain sensitive prompts and personal data?Next in this trackHow would you detect someone stealing your model through its API?Next in this trackDesign LLM-assisted phishing triage for a security team, and list its failure modes.
DISCUSSION · 0

No comments yet. Share a correction or a different way to answer it.