The best answer is C. Determine the business objective and stakeholder needs . In PMI-CPMAI, the first work in the business understanding phase is to clarify the problem to be solved, define the business question, and align the initiative to stakeholder expectations before deciding whether AI is necessary or which parts are cognitive. PMI’s official CPMAI exam outline explicitly includes formulating AI-specific business questions, prioritizing and scoping AI projects, conducting go/no-go assessments, and separating cognitive from non-cognitive components. That sequence strongly suggests that the team must first understand the business objective and stakeholder needs before evaluating alternatives, transparency controls, or data requirements.

Option B is important, but it comes after the project team understands what the business is trying to achieve. Only then can they assess whether non-cognitive alternatives are sufficient. Option A is also relevant, especially for trustworthy AI, but ethical and transparency requirements should be defined in the context of the business use case. Option D is premature because data identification follows the initial business framing. PMI’s methodology is built around starting with business understanding, then deciding whether AI is appropriate, which pattern applies, and what governance is required.