Org-level
Decisions that are made on an organizational level and/or impact multiple departments
Driver runs it · Approver signs off · Contributors support execution · Informed are updated
Org direction & AI ambition
Adds a feasibility and sequencing read to the ambition the executive team sets.
Role & performance expectations for AI use
Helps define what good AI use looks like by role. People owns the expectation and the performance conversation.
Org-wide fluency floor
Advises on level and content. Delivering fluency training at scale is a separate skill set, and usually sits better with People/L&D or an external partner.
Cross-functional AI opportunity roadmap
Owns prioritization by ROI, feasibility, and strategic impact. Depends on functional leaders being equipped to contribute use cases from their own function.
Platform & approved tool standard
Owns the recommendation and the migration. Consolidation is what makes shared skills libraries and shared projects possible.
Shared reference architecture & build standards
Retrieval patterns, agent design, system-of-record connectivity, and the standards for monitoring and evaluation.
Org context layer & shared data infrastructure
The connective tissue that makes company knowledge queryable. Data cleanliness and legibility usually surface here first.
Security, data handling & AI risk policy
Co-owns technical standards — risk register, audit trails, acceptable use. Policy decisions sit with Security and Legal.
Resource & token management
Supplies usage patterns. One person tracks token use across the org and shares the data back to teams, similar to how a controller tracks spend.
Champions program: whether, scope, mandate
Shapes the scope and later coaches the champions. The mandate and capacity plan come from the executive sponsor.
Measurement approach by phase
Supplies what's measurable at each stage, so early efforts aren't judged against later-stage metrics.