Learning Objectives

  1. Enumerate prompt-injection and jailbreak categories (direct, indirect, multi-turn, role-play bypass) and mitigations
  2. Build a PII redaction pipeline that sanitizes LLM inputs and outputs (names, emails, SSNs, PHI)
  3. Apply log sanitization to remove sensitive data from AI system audit logs
  4. Complete a NIST AI RMF 1.0 mapping worksheet (Govern, Map, Measure, Manage) for a real AI deployment
  5. Relate NIST AI 600-1 GenAI Profile controls to the lab scenarios
  6. Describe how EU AI Act risk tiers apply to AI-assisted security tools (high-risk deadline: 2 Dec 2027)

Lecture Notes

Opens Wednesday morning

The objectives above are the full picture of what Day 3 covers. Lecture notes, slides and lab guides are released when we get there, so the room works through them together.

Lab Guide & Bundle

4 labs, released Wednesday morning

Lab guides and downloadable bundles appear here when Day 3 opens.

End-of-Day Quiz

Check for understanding — reveal the answer after you've chosen. No score is recorded.

  1. Data can be protected in three states. Which control specifically protects data 'in use' — while it is being actively processed in memory by the model?

    • TLS / HTTPS
    • Full-disk or database encryption at rest
    • Confidential computing / trusted execution environments (encryption in use)
    • A daily backup schedule
    Reveal answer

    Correct: C. Confidential computing / trusted execution environments (encryption in use)

    Encryption in transit (TLS) protects data on the wire; encryption at rest protects stored data; encryption in use — confidential computing / TEEs — protects data while it is decrypted and being processed. AI inference exposes the 'in use' state, which is why it gets special attention.

  2. In the Day 3 lab, an HR assistant holds employee SSNs in its context and attackers try to exfiltrate them. Which control most reliably eliminates the data-exfil attack surface?

    • Asking the model politely not to reveal PII
    • Redacting/anonymizing PII at the input layer with Presidio before it ever enters the context
    • Making the system prompt longer
    • Relying on the model's alignment training to refuse
    Reveal answer

    Correct: B. Redacting/anonymizing PII at the input layer with Presidio before it ever enters the context

    You cannot leak what is not there. Redacting PII at the input (Microsoft Presidio detects PERSON, US_SSN, EMAIL, etc. and replaces it with placeholders) removes the sensitive data from the context entirely. Model refusal and alignment are unreliable and must not be treated as a data control.

  3. NIST AI RMF 1.0 is organized around four core functions. Which set is correct?

    • Identify, Protect, Detect, Respond
    • Govern, Map, Measure, Manage
    • Plan, Do, Check, Act
    • Collect, Train, Deploy, Retire
    Reveal answer

    Correct: B. Govern, Map, Measure, Manage

    NIST AI RMF 1.0 uses Govern, Map, Measure, Manage. (Identify/Protect/Detect/Respond/Recover is the NIST Cybersecurity Framework — a common distractor.) The lab worksheet maps the HR-chatbot scenario to these four functions; NIST AI 600-1 adds a Generative AI Profile of controls on top.

  4. Under the EU AI Act, what is the provisional (Digital Omnibus, pending adoption) deadline being tracked for high-risk AI system obligations in this course?

    • 2 December 2025
    • 2 December 2026
    • 2 December 2027 (deferred, provisional)
    • There is no deadline — the Act is voluntary
    Reveal answer

    Correct: C. 2 December 2027 (deferred, provisional)

    The course tracks the deferred high-risk obligations date of 2 December 2027 (provisional, per the Digital Omnibus proposal, pending final adoption). Always flag it as provisional and re-verify at content lock, because EU AI Act timelines have shifted.

  5. 'Shadow AI' in an organization refers to what?

    • AI models that only run at night to save power
    • Employees using unsanctioned AI tools/services outside governance, risking data leakage and compliance gaps
    • A backup copy of a production model
    • An adversarial model trained to attack your model
    Reveal answer

    Correct: B. Employees using unsanctioned AI tools/services outside governance, risking data leakage and compliance gaps

    Shadow AI is the AI analogue of shadow IT: staff pasting sensitive data into unapproved chatbots or standing up ungoverned AI features. It bypasses data-protection and responsible-AI controls, which is why discovery and policy (a Govern-function concern) matter.