Aerial view of Boatwright Tower

AI & University Policy

October 1, 2026

Artificial intelligence creates the potential for increased efficiency and new capabilities for work at the University of Richmond. At the same time, the security and integrity of the information and processes of the University must be ensured. While there are several policies for the security of administrative data, two stand out when it comes to artificial intelligence: the Administrative Data Management Policy and the Data Security Policy.

Administrative Data Management Policy

While it’s recommended to read the full policy, what’s most pertinent for AI users are the notions of the Administrative Data and Data Stewardship Domains, and the Data Stewards themselves. 

Administrative Data is defined as "Data that is gathered, produced, stored, and/or disseminated concerning any aspect of the University’s operations. Such data includes, but is not limited to, general ledger/accounting, human capital management, student, alumni/development, space management, student housing data, and also includes any data derived from the use of such data. Administrative data specifically does not include faculty generated classroom data nor faculty- or student-generated research data."

Data Stewardship Domains are "A University administrative or operational area, function or process and the authoritative data elements created, collected, and managed by that area throughout the lifecycle of that data, regardless of the Transactional System(s) in which they reside (e.g., Finance data)."

And a Data Steward is "typically a senior operational manager in a functional area responsible for a Data Stewardship Domain." Data Stewards are responsible for maintaining the security and integrity of the data in their domains. The policy lists every Data Stewardship Domain and its Data Stewards in a useful appendix.

Data Security Policy

The Data Security Policy is also worth reading in its entirety. For staff and faculty who want to use University Administrative Data with an AI, there is a whole section of the policy that spells out the requirements. In brief, these requirements state: 

  • University Administrative Data may only be entered into systems that have been approved. 
  • Each generative AI system must have an Acceptable Use Policy, which includes user responsibilities. Users of AI systems should be aware of their responsibilities. 
  • Entry of Administrative Data into an AI must be approved by a Data Steward for each type of Data Stewardship Domain. 
  • Any output from an AI must be reviewed by the appropriate Data Stewards before it can be entered into another University system. They will review the information for confidentiality, integrity, accuracy, fairness, regulatory compliance, and academic attribution. 
  • When an AI provides data or a decision, a disclaimer must clearly state that AI was used to generate the output. This ensures transparency and informs users about the nature of the information they are receiving. 

Artificial intelligence technology continues to make advances quickly. As it does, the University continues to refine its policies around the use of University Administrative Data. This fall, the AI Coordinating Council has launched a Usage Guidelines Working Group, led by Melissa Hartz, executive director of the Office of Institutional Effectiveness. Among the group’s responsibilities are updates to University policies. If changes are made to policies, those changes will be communicated to the University community.