Question 610
A reviewer challenges the choice of AI training dataset access minimization when two research groups must train on separate approved subsets of a sensitive dataset. Which underlying operation justifies the choice?
Explanation
Correct Answer
The operative mechanism for AI training dataset access minimization: Limit training service accounts to designated sanitized data resources with scoped storage and dataset grants. For this scenario (two research groups must train on separate approved subsets of a sensitive dataset), verify that each research group accesses only its allocated training slice in acceptance tests.
Incorrect Answers
This describes how Gemini Agent Platform Agent Identity works instead: Use Agent Identity and appropriately scoped policy grants for each agent workload and tool-access path. It is useful when a support agent should not inherit broad project editor permissions simply because its developer has them, but it does not explain the selected AI training dataset access minimization for the current requirement (two research groups must train on separate approved subsets of a sensitive dataset).
This describes how Gemini Agent Platform security findings works instead: Use supported Agent Platform Security Command Center integration and Cloud Audit Logs to review agent-related findings and attributable policy changes. It is useful when security operations wants to correlate risky tool activity with agent platform audit events, but it does not explain the selected AI training dataset access minimization for the current requirement (two research groups must train on separate approved subsets of a sensitive dataset).
This describes how Gemini Agent Gateway policy enforcement works instead: Configure the supported Agent Gateway path with enforced runtime access and policy controls. It is useful when a customer support agent requires one controlled gateway for all approved model and tool interactions, but it does not explain the selected AI training dataset access minimization for the current requirement (two research groups must train on separate approved subsets of a sensitive dataset).






Professional Cloud Security Engineer Practice Tests