Applied AI
at EWU

What does the institution need to decide next—and what evidence is still missing?

The assignment began as a preliminary scan of low-risk applied AI use cases. The work expanded into an evidence sequence: current activity, external context, a faculty-facing catalog, risks and open questions, practical guidance, a supporting framework, and a final decision document.

AI activity is already underway. The institutional problem is how to connect evidence, responsibility, support, and governance.

Connect live practice to the decisions that can responsibly support it.

The research distinguishes AI literacy, applied AI, and AI science; observed activity from predicted use; faculty work from enterprise administration; and evidence from recommendations. The five-condition low-risk test creates a repeatable boundary around data, consequential decisions, human accountability, policy fit, and disclosure.

Fourteen of nineteen defined faculty use cases meet that test with named conditions. Five remain not yet classifiable because a data or external-policy condition is unresolved.

The deliverables form a sequence, not a document stack.

• R1Preliminary Scan: campus evidence, faculty use catalog, and external landscapeDelivered 30 Jun
• R2Problems, Risks & Open Questions Brief with five-condition low-risk testDelivered 20 Jul
• R3AI at EWU guide and Faculty AI ToolkitInternal working draft / 05 Aug
• R4Final AI Report: seven decisions, two capacity directions, and handoffDelivered 19 Aug
• SupportApplied AI Framework DraftStandalone

Each stage narrowed uncertainty before making recommendations.

1. Map activity and context. Combine a faculty survey, known campus practice, public tool baselines, regional peers, and national anchors.
2. Name evidence gaps and risks. Keep prospective risks unranked when the institution lacks evidence to rank them responsibly.
3. Define low-risk conjunctively. A use must satisfy all five conditions; uncertainty on one condition means “not yet classifiable.”
4. End with ownership and decisions. The final report frames seven decisions and two directions for durable institutional capacity rather than presenting one predetermined structure.

The summer work is delivered. The institutional work is not.

The report transfers the next action to governance and institutional owners: decide shared definitions, curricular direction, coordination, source-of-truth maintenance, support capacity, and the route for unresolved risks.

A continuing owner for the faculty guide was not named in the supplied materials at delivery. The guide therefore identifies a future review window and treats source offices as authoritative for their own material.