The correct option is AI accountability and explainability.

The scenario emphasizes that executives want clear ownership of each recommendation and that auditors must be able to understand how the assistant produced its suggestions. This is exactly what this priority addresses because accountability establishes who is responsible for model outputs and operational decisions, and explainability provides human understandable insights into how models arrive at their predictions through techniques such as feature attributions and transparent logs.

Under tight regulatory oversight, organizations need audit trails, traceability, and mechanisms to justify outcomes. This priority aligns with governance practices on Google Cloud such as using audit logs to track who did what and when, along with explainability tools in Vertex AI that help interpret model behavior so that stakeholders can trust and verify the assistant’s recommendations.

Data storage costs and efficiency is incorrect because the scenario is not about optimizing storage spend or capacity. It is focused on governance and transparency for model decisions.

Vertex AI Model Garden is incorrect because it is a catalog of models and solutions. It does not define a governance focus on who is responsible for recommendations or how predictions are explained.

Model inference speed and latency is incorrect because performance tuning for response time does not address the need for accountability or the ability for auditors to understand model reasoning.

Full AWS Practitioner Certification Question