Human-in-the-Loop and Accountability

1. What does the 'Human-in-the-loop' approach primarily ensure in AI governance?

Decentralized decision-making Automatic AI outputs Named person accountable for AI decisions Reduction of AI complexity

2. What is the key requirement for AI outputs that are considered high risk according to the lesson?

They must require no human intervention They must be generated quickly They require mandatory human review They can be approved by algorithms only

3. How does LyraLearn implement accountability in its AI processes?

By allowing AI-generated content to bypass human reviews By maintaining an append-only audit log of approvals and rejections By employing only automated systems for review By ignoring low-stakes outputs

4. What is an escalation path in the context of Human-in-the-loop design?

A way to approve outputs quickly A method for escalating uncertain outputs to senior reviewers An automatic rejection process A direct approval from the AI model

Continue: Mapping to Compliance → Review this lesson Retake quiz