Students
Provide clear approved tools, literacy and boundaries for coursework and institutional information.
University AI Governance
One university. Different AI environments.
A first-year student, a research commercialization office, a clinical researcher and an administrative team do not necessarily need the same AI environment, permissions or governance.
Different responsibilities
Route AI by user, information and purpose
University governance becomes more practical when it distinguishes real operating contexts.
Faculty
Support teaching and scholarship while respecting copyright, privacy, assessment and institutional policy.
Researchers
Protect unpublished findings, datasets, methods, sponsored work and emerging intellectual property.
Administration
Review permissions, records, confidential information and approved enterprise tools.
Commercialization
Protect patent timing, confidential disclosures, licensing opportunities and industry agreements.
Sensitive research
Use stronger controls where clinical, regulated, security-sensitive or contractually restricted information is involved.
Shadow AI
Policy works better when there is an approved alternative
Students, researchers and employees will naturally use capable tools that help them work. If approved institutional AI is too limited or unclear, important information can move into unmanaged services through ordinary behaviour.
AFA helps institutions combine policy with useful approved environments, training, workload routing and clear escalation.
Governance path
Classify → Approve → Enable → Review
Governance should make the right behaviour easier to follow.
Classify
Define information and workload categories.
Approve
Map categories to tools and environments.
Enable
Train each community according to its responsibilities.
Review
Monitor changes in tools, models, policy and institutional needs.
Continue the journey