A consultant is preparing to roll out a new generative AI feature in Dynamics 365 Sales to a 400-seat sales organization. Before enabling the feature tenant-wide, the consultant needs to confirm the organization will not hit avoidable rollout problems in week one. (Select TWO.)

The correct choices are the confirmation of sufficient AI capacity or credit licensing for the planned seat count and the confirmation that sellers' security roles grant access to the underlying Dataverse tables and fields the AI feature needs to read. A generative AI feature can only return complete, accurate results if the sellers using it hold security roles with access to the Dataverse tables and fields the feature reads from, so confirming that access before go-live prevents silent failures or incomplete answers in week one. Separately, generative AI features consume message or credit-based capacity, and a 400-seat rollout can exhaust an underprovisioned capacity allotment almost immediately, so confirming licensing capacity matches the planned seat count prevents the feature from failing or throttling for sellers shortly after launch. The multi-day AI certification course requirement is incorrect because there is no requirement for sellers to complete a multi-day certification course before using the feature; light onboarding may help adoption, but it is not a technical or licensing prerequisite that would cause week-one failures if skipped. The localization into every customer-facing language requirement is incorrect because localizing the feature into every customer-facing language is not required for the feature to function correctly for internal sellers, and skipping it would not cause the kind of rollout failure the consultant is trying to avoid. The two real prerequisites here are data access and capacity, both of which directly determine whether the feature works at all once turned on.