RAIUC02-BP02 Identify contributing and other upstream stakeholders

Map out the roles involved in your AI system's lifecycle from initial concept to ongoing operations. Consider product, engineering, data science, legal, security, infrastructure, and other company functions that provide input on requirements or constraints.

Identify external stakeholders who contribute to or influence your system even though they are not part of your direct team. This includes vendors who supply data or model components and upstream teams whose decisions affect your system's design or operation.

Document the specific ways each stakeholder group contributes to or influences your system, rather than just listing names and titles. For example, note that your security team reviews threat models and sets deployment constraints, while your legal team provides guidance on data use and compliance obligations, and your infrastructure team manages the computing Resources your system runs on.

Include oversight and governance stakeholders who may not be involved in day-to-day development but have authority over key decisions about your system. This covers compliance officers who approve deployments, risk management teams who set acceptable use policies, and executive sponsors who control funding and strategic direction.

Create a stakeholder contact list with clear points of contact for each group, including backup contacts and escalation paths for important decisions. Keep this list updated as team structures change and make sure everyone on your team knows who to reach out to for different types of questions or approvals throughout the system's development and operation.