AAIA: Advanced in AI Audit

By Isaca · Question Mode

QuestionQ57

AI Governance and Risk

A financial institution’s customer-service chatbot occasionally provides users with incorrect legal advice, resulting in several customer complaints and possible regulatory exposure. Which of the following should be done FIRST to mitigate this risk?

Explanation

A human-in-the-loop review process provides an immediate control over high-risk legal guidance by ensuring qualified people can review, correct, or withhold unsafe chatbot responses before they affect customers. Dataset and model improvements may help over time, but do not by themselves provide the same immediate safeguard.

Community Discussion

No comments yet. Be the first to start the discussion!