1. A requirement reads "The system shall respond optimally under all conditions." Which risk does this illustrate?
2. Which statement correctly distinguishes outdated knowledge from model drift?
3. An AI generates job applicant personas that consistently portray leadership candidates as male, aged 35 to 50, with a finance background. Which risk is this, and why does it matter in RE specifically?
4. Why does the syllabus warn that anonymised data may still create a data leakage risk?
5. For text output, when does the EU AI Act labelling obligation described in the syllabus apply?
6. Under what condition is disclosure of AI-generated content not required?
7. Which risks belong to the group arising from how AI is used within a process, rather than from the model itself? Select all that apply.
8. What makes context window overflow particularly difficult to guard against?
9. A team adopts AI-generated safety requirements without proper human review, and auditors later find contradictions with mandatory standards. Which risk does this illustrate?
10. Why does opacity complicate requirements validation, and not only stakeholder trust?