QuestionQ31

Securing AI systems

A data-science team is prepared to release a model for enterprise use. The team wants to safeguard the model against accidental changes or tampering. Which of the following actions is most appropriate?

Explanation

An API integrated with IAM roles enforces authenticated, role-based, least-privilege access to the model. This limits model interaction and administrative actions to authorized identities, helping prevent unauthorized or unintended changes and tampering.

Community Discussion

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