A governance lead is designing a checklist for new AI projects and wants exactly one question per pillar. Which question belongs under 'accountability'?
Select an answer to reveal the explanation.
Short Explanation
Accountability is the ownership question. You need a named owner who’s responsible for the model’s decisions and steps in when something goes wrong. Explainability, robustness, and fairness each ask something different.
Full Explanation
The correct choice is the question asking who is the named owner responsible for the model's decisions and for acting when something goes wrong. Accountability is specifically about assigning a responsible owner for a model's outcomes and remediation. The option asking whether an affected user can understand why the model produced a particular result is incorrect because it asks about explainability. The option asking whether the model maintains acceptable performance when inputs are noisy or unexpected is incorrect because it asks about robustness. The option asking whether outcomes are reasonably consistent across different demographic groups is incorrect because it asks about fairness. For the exam, be able to write or recognise a checklist item that maps cleanly to each of the five pillars.