Skip to content
AI Security

Interview question · Prompt Injection & Jailbreaks

Walk through a zero-click indirect injection that ends in data exfiltration, and the controls that would have broken the chain.

Use the historical EchoLeak advisory and an original worked trace to distinguish delivery, actual exposure, generated output and release. Choose verifiable controls without implying the old vulnerability remains open.

hardNewUpdated Nov 2026

Use the historical EchoLeak advisory and an original worked trace to distinguish delivery, actual exposure, generated output and release. Choose verifiable controls without implying the old vulnerability remains open.

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.

signing in unlocks more free answers per topic

What if · the follow-ups in this answer

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

Hana requests a staff project update from a fictional assistant. Accepted external mail can be retrieved alongside permitted internal notes; an answer-preview component can request remote resources.

  1. What if the external message was delivered but never included in a model request?
  2. What if the browser blocks the resource but a backend preview service fetches it?
  3. What if the fixture shows a correct answer and a denied preview request?
  4. What would a responsible incident handoff include?

Each rung has a nudge, a model answer, the principle that carries over and the weak answer to avoid. Sign in free to open it.

UP NEXT ON YOUR JOURNEY
Next in this trackDo delimiters, spotlighting and an instruction hierarchy help against prompt injection? What can you honestly claim for them?Next in this trackHow has the move from chatbots to agents changed the impact of prompt injection?Next in this trackHow do you evaluate many-shot and multi-turn jailbreak defenses?
DISCUSSION · 0

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