Question 18
Domain 6: Human-AI Interaction and OversightWhat accountability framework would best ensure appropriate human oversight?
Correct answer: B
Explanation
Appropriate human oversight requires humans to remain in control of high-risk decisions, so “human approval gates” and an “override capability” ensure AI actions can be stopped or modified. “Explanation provision,” “training program,” and “accountability documentation” support informed supervision and an audit trail showing that humans reviewed and monitored the system.
Why each option is right or wrong
A. Let AI make all decisions autonomously with no human involvement.
B. Implement accountability framework with: human approval gates (humans must approve high-risk trades), override capability (humans can stop or modify trades), explanation provision (clear reasoning for each decision), training program (teach humans to supervise AI), and accountability documentation (audit trail shows human oversight).
The selected framework is the only one that operationalizes human oversight at the decision point: for high-risk trades, a human approval gate and a live override function ensure the system cannot execute without review and can be stopped or modified before harm occurs. The explanation, training, and documentation elements are the supporting controls that make oversight meaningful and auditable, creating a traceable record that a human actually supervised the AI rather than merely being nominally responsible.
C. Have humans make all decisions without AI assistance.
D. Require humans to review decisions after execution.