A library consortium wants consistent AI vendor reviews across branches. Which approach best standardizes supply-chain due diligence?

A library consortium wanting consistent AI vendor reviews across branches should use a shared vendor AI questionnaire covering retention, data residency, and training opt-out. That approach best standardizes supply-chain due diligence because a common core lets every branch evaluate the same risk dimensions, compare answers, and avoid gaps where one site forgets residency or training-use questions. Shared questionnaires also accelerate consortium-wide decisions and reduce contradictory local practices.

Using a shared vendor AI questionnaire covering pricing tiers, uptime SLAs, and support hours fails conceptually because it standardizes commercial terms rather than AI risk dimensions, so branches still won’t evaluate retention, data residency, or training opt-out consistently. Using a shared vendor AI questionnaire covering benchmark scores, latency metrics, and feature depth fails because model performance and features don’t reveal vendor data retention, residency, or AI security posture. Using a shared vendor AI questionnaire covering prompt libraries, branch workflows, and training guides fails because it standardizes usage enablement rather than AI risk dimensions, addressing operational adoption, not pre-award AI supply-chain due diligence.

Exam caveat: a shared questionnaire still needs local supplements for unique branch systems; the common core is minimum assurance, not a ban on justified extras. Operational check: publish the consortium AI questionnaire, require completed responses before branch pilots, centralize scored results, and escalate incomplete retention/residency/opt-out answers as non-award conditions.