Question 970
Following an approved migration, the team revisits: A company wants to satisfy NIST AI risk governance but measures only firewall block counts. What test result is most decisive?
Explanation
Correct Answer
Governance evidence maps major AI risks to accountable owners and measured mitigation tests. It validates the recommended action: Create an AI risk register and review model, application, data and human-oversight controls against the relevant framework. The design principle is: Security telemetry alone does not demonstrate governance, accountability or model oversight.
Incorrect Answers
Only approved AI tools meet the user-group and data-handling policy; unknown AI tools are monitored or blocked. That observation supports a DIFFERENT situation: A team allows all AI writing tools because their URLs appear in the productivity category. Here the selected design is: Create an AI risk register and review model, application, data and human-oversight controls against the relevant framework.
The summary task succeeds while a test purchase approval action without human authorization fails. That observation supports a DIFFERENT situation: An internal AI procurement assistant is approved for supplier summaries but begins requesting purchase approval tokens. Here the selected design is: Create an AI risk register and review model, application, data and human-oversight controls against the relevant framework.
Live prompts produce runtime enforcement logs; model scans remain associated with immutable artifact versions. That observation supports a DIFFERENT situation: A developer proposes using AI model scanning to protect every runtime response, without request-path integration. Here the selected design is: Create an AI risk register and review model, application, data and human-oversight controls against the relevant framework.