MS-CC Guidance for AI Policy Development

This document is designed as a flexible starting point, not a prescriptive mandate. Institutions at any stage of AI policy development — whether just beginning or already iterating on existing frameworks — will find relevant guidance and recommended areas of action tailored to the MSI context.

The MS-CC AI Policy Guidance was developed through a rigorous, community-driven process drawing on national best practices, peer institution expertise, and direct input from practitioners across HBCUs, TCUs, and Minority-Serving Institutions.

Landscape Review

The MS-CC team reviewed existing national and regional conversations on AI use in higher education — drawing from attendance at national and sector-specific events, published literature and policy frameworks, and insights from peer institutions and professional networks engaged in early AI adoption

Development of the AI Policy Shell

From the landscape review, the MS-CC team synthesized findings into a structured document of prompting questions organized into distinct subsections, designed to serve as a conversational tool for institutions rather than a prescriptive policy.

Structured Working Sessions

The MS-CC consolidated its Community of Practice sessions for CI Planning, Cybersecurity, and IT Leaders into a single focused session, in collaboration with AIHEC. The AIHEC AI Working Group — which recently developed a document outlining AI policy guidelines for Tribal Colleges and Universities — was invited to present its guidelines and participate in the working session. Engagement and feedback were received asynchronously and across two facilitated sessions, ensuring the guidance was grounded in the practical realities faced by HBCUs, TCUs, and similar institutions.

Synthesis into Policy Guidance

Common themes were elevated as recommended areas of consideration. Unique institutional examples were preserved to reflect the variety of contexts, and emerging gaps were documented for future exploration.

The following themes emerged consistently across working sessions and reflect the most pressing considerations for MSIs developing or refining AI policy. These takeaways informed the structure and content of this guidance document.

Acceptable Use & Accessibility

Institutions need clear, enforceable policies that define acceptable AI use and ensure equitable accessibility across all campus constituents, including those with limited technical backgrounds or requiring accommodations.

Opt-Out Mechanisms

Community members must have meaningful pathways to opt out of AI tools — especially in shared meeting or classroom settings. Policy should clearly define when and how opt-out is available and what alternative workflows exist.

AI Infrastructure Readiness

Standing up AI infrastructure requires significant planning around compute resources, dedicated CPU/GPU capacity, interlinks between machines, and stateful environment management for training and inference workloads.

Storage & Agentic Workflows

Storage policies must address the unique demands of agentic AI workflows, including data lifecycle management, segregation by sensitivity, and secure decommissioning of datasets and models at end of life.

Training at All Levels

Effective AI governance requires training spanning faculty, staff, students, and leadership — addressing not just how to use AI tools, but where training data comes from and what biases may be embedded in the models.

Vendor LLM Integrations

Vendors are actively embedding LLMs into existing products and platforms. Institutions need structured processes for evaluating what AI capabilities vendors are introducing and whether they meet institutional data governance standards.

Internal vs. External Tool Barriers

Barriers to ML and LLM adoption often stem from unresolved policy questions around reliance on internal vs. external tools and how data sovereignty is maintained when using third-party platforms.

Communities of Practice

Group forums, communities of practice, and cross-functional working groups are essential infrastructure for enabling ongoing AI learning, policy iteration, and institutional knowledge sharing across campus.