AI AI in Teaching and Learning

Policy artifacts and why they matter for teaching

Campus AI governance produces many documents. Here is what each one is for, who reads it, and how it shapes the classroom.

Key takeaways
  • AI governance produces documents at seven layers, from principles to evidence.
  • Every layer affects the classroom, even those instructors never read.
  • Visibility labels decide who can see each document.

Seven documentation layers

Each layer is narrower in scope and less widely published than the one above. All of them affect the classroom, including the ones instructors never read.

1. Institutional AI governance Strategy, principles, acceptable use, data classification, academic integrity guidance and the consolidated risk registry. Teaching & learning: Sets the campus rules, such as integrity and data limits, that every syllabus inherits.
2. NIST AI RMF profiles Context profiles (teaching and learning, research, administration) and platform profiles built on Govern, Map, Measure, Manage. Teaching & learning: A teaching and learning profile frames the risks of classroom AI, from AI feedback to student data in course tools.
3. Service catalog & use-case registry What AI services exist, and which uses are approved or proposed. Teaching & learning: Shows which tools and teaching uses are already approved, so instructors adopt vetted practices.
4. AI service cards User-facing documentation: what a tool is, for whom, and under what rules. Teaching & learning: The rules students and instructors follow when they use a specific tool for coursework.
5. System cards Deployment-level architecture, integrations, controls and governance boundaries. Teaching & learning: Explains what happens to student data when AI is built into the LMS or other course tools.
6. Model cards Underlying model capabilities, limits and evaluation context. Teaching & learning: Gives facts about accuracy, bias and limits to use in AI-literacy lessons.
7. Supporting evidence Risk registers, reviews, data-flow diagrams, evaluations and incident records. Teaching & learning: The privacy and accessibility reviews that make a tool safe to require in a course.

Artifact types

Each artifact answers a different governance question. Every card notes how it affects teaching. Cards marked “Used directly in courses” are ones faculty and students read themselves.

Governance Public

Institutional AI principles

State institutional posture and values.

In teaching & learning Sets the values course policies inherit, such as AI literacy, human judgment and integrity. Instructors can cite them in a syllabus.

For: Community, leadership, external stakeholders

Governance Public / Institution-wide Used directly in courses

Acceptable use guidance

Define allowed and prohibited use.

In teaching & learning The campus baseline for students and faculty. Course policies tighten or loosen AI use within it.

For: Students, faculty, staff, researchers

Governance Public / Institution-wide Used directly in courses

Academic integrity & teaching guidance

Explain how AI relates to integrity and teaching practice.

In teaching & learning Sets the default rule when a syllabus is silent, disclosure expectations and how suspected misuse is handled. Often the source of sample syllabus statements.

For: Students, faculty, teaching support

Profile Restricted Internal

AI RMF profiles

Adapt risk management to specific contexts.

In teaching & learning A teaching and learning profile sets expectations for AI tutors, AI-assisted feedback or grading, and student data in course tools.

For: Governance teams, service owners, reviewers

Catalog Public / Institution-wide Used directly in courses

AI service catalog

List supported tools and entry points.

In teaching & learning Shows which tools students can be asked to use at no cost to them, which supports equitable access.

For: Entire campus

Catalog Institution-wide / Restricted

AI use-case registry

Index approved and proposed uses, mapped to context profiles.

In teaching & learning Records approved teaching uses, such as AI feedback on drafts or a course tutor, so instructors build on vetted practices.

For: Service owners, approvers, governance teams

Service Public / Institution-wide Used directly in courses

AI service cards

Explain what a tool is, for whom, and under what rules.

In teaching & learning Tells students what data they can enter into a course tool, and gives instructors rules to cite in assignment instructions.

For: End users, approvers, support teams

System Restricted Internal

System cards

Describe the deployed institutional system.

In teaching & learning Matters when AI is integrated into the LMS or courseware. It documents data flows, logging and controls behind student-facing features.

For: Security, privacy, technical owners, auditors

Model Public (vendor) / Restricted (internal)

Model cards / summaries

Describe underlying model capabilities and limits.

In teaching & learning Helps instructors explain a tool’s limits and choose models suited to their discipline.

For: Reviewers, developers, governance teams

Evidence Confidential / Restricted

Risk and review evidence

Support assurance, approval and auditability.

In teaching & learning Privacy (FERPA) and accessibility reviews behind course tools. Instructors rarely see them but rely on them when adopting a tool.

For: Internal reviewers and leadership

Four visibility labels

Classroom guidance should be easy to find. The reviews behind it usually stay internal.

Public Safe for open publication, including prospective students and the open web.
Institution-wide Visible to all campus users with institutional credentials, but not published openly.
Restricted Internal Limited to approved internal stakeholders such as governance bodies and service owners.
Confidential Limited due to sensitive security, privacy, legal or incident content.

Six instruments, six jobs

Name which kind of document you are writing. Each does a different job.

Policy Sets what is required. Students must disclose AI use in assessed work.
Guidelines Explain how to comply. How to write an AI-use note.
Procedures Define how the work gets done. Steps for reviewing a suspected integrity case.
Standards Set consistent requirements. Approved tools must pass privacy and accessibility review.
Learning resources Build capability. A workshop on designing AI-aware assignments.
Assurance Show compliance and effectiveness. An annual review of integrity cases and student survey results.

An operating model for classroom AI

Policy levels need people and processes behind them. Each area below needs a named owner.

Governance and accountability Who sets expectations, approves exceptions, and owns academic integrity, privacy, security and risk?
Technology and infrastructure Which tools are available, and how are they evaluated, bought, integrated, secured and supported?
Faculty development How do instructors build AI literacy, teaching practice and assessment-design skill?
Student enablement How do students build AI literacy, disciplinary competence, critical thinking and responsible practice?
Measurement and improvement How does the institution track learning, student experience, equitable access and whether policies work?

Layers, artifacts and labels are from the Campus AI Registry & Governance Framework (artifacts, publication review). The teaching and learning notes are our own applications.

Disclosure

The Campus AI Registry & Governance Framework is also developed by Joe Sabado, the author of this guide. It is an original framework, not an established external standard or a documented practice at any institution. It builds on the NIST AI Risk Management Framework, which is an external standard.