V1203.03 Artificial Intelligence Policy

A. Purpose

Valley City State University (VCSU) recognizes the transformative potential of artificial intelligence (AI) in advancing education, research, and innovation. This policy establishes a framework for the responsible and ethical development, deployment, and use of AI technologies within the university community.

NOTE: Sections A and C were partially written with ChatGPT as a writing tool.

B. Scope

This policy applies to all employees and students at Valley City State University.

C. Definitions

AI System: A computer-based system that utilizes algorithms, machine learning, or other techniques to perform tasks that typically require human intelligence, such as problem-solving, decision-making, language understanding, and pattern recognition.
AI Project: An AI project is a comprehensive undertaking that aims to develop, implement, or improve an AI system. It involves multiple components, such as problem definition, data collection, algorithm development, model training, testing, deployment, and ongoing maintenance.
AI Query: An AI query typically refers to a single interaction or request made to an existing AI system. It involves posing a question or providing input to obtain a specific output or response. AI queries are generally straightforward and do not encompass the entire process of developing or implementing an AI system.

D. Principles:

Ethical Considerations: VCSU commits to ensuring that AI technologies are developed and used ethically, prioritizing fairness, transparency, and accountability. Each user will scrutinize results obtained from AI queries they conduct. AI project managers and researchers shall avoid creating or deploying AI projects that may contribute to discrimination, bias, or harm to individuals or groups. While the VCSU AI Review Board provides some oversight and will act on complaints and grievances, it is ultimately the responsibility of the individual user, AI project manager or researcher to decide if the use and output of AI in a particular instance is ethical.
Academic Freedom and Collaboration: VCSU encourages open and collaborative application of AI. Employees, researchers, and students are free to explore and innovate within the bounds of this policy, fostering a culture of academic freedom while respecting legal, policy and ethical constraints.
Data Privacy and Data Security: VCSU is committed to protecting the privacy of individuals in the collection, storage, and use of data for AI applications. Data used with AI systems shall be responsibly obtained and shall comply with applicable data protection laws, regulations, and policies, including NDUS Procedure 1203.7 “Data Classification and Information Security Standard”.
Copyright: Users shall recognize that query output from AI systems may be protected by copyright laws or involve intellectual property rights.
Transparency: AI systems developed at VCSU should be transparent and explainable, including origins of data, algorithms and techniques used. The AI project manager or researcher must always be prepared to provide understandable explanations for the decisions made by AI systems, particularly when they impact individuals.
Inclusive Access: VCSU aims to make AI education and resources accessible to all members of the university community. Efforts will be made to minimize barriers and provide support for individuals from diverse backgrounds and abilities to engage with AI technologies.
Research and Application Excellence: VCSU encourages AI research and application that contributes to societal well-being. Users involved in AI projects shall prioritize human beings, and design and test to ensure that systems are safe and reliable and do not pose any unnecessary risks to users or society.

E. Governance, Oversight and Process:

1. AI Review Board:

An AI Review Board oversees the development and deployment of AI Projects and Systems at VCSU. Furthermore, the AI Review Board will review associated complaints or grievances.
The AI Review Board will consist of the Chief Information Officer, one Technology Services staff member appointed by the Chief Information Officer, the Chair of the IRB (Institutional Review Board), the Chair of the CSSE (Computer Systems and Software Engineering) department or their designee, and one faculty member selected by the Faculty Association and serving a two-year term.

2. Compliance:

All VCSU AI projects must undergo a review to ensure compliance with this policy. There is one exemption to this requirement: A student created AI project as part of a course requirement is exempt, provided all of the following requirements are met: a) the AI project does not include information classified as private, restricted or confidential by policy, procedure or law; b) the project will not result in an AI system used for official VCSU business; c) the project will not be published or presented publicly; d) the course instructor provides guidance and oversight of the student AI project. The AI Review Board will have the authority to halt or modify projects that do not meet the principles or standards outlined in this policy. Scheduled periodic reviews may be required for selected significant projects or systems at the discretion of the AI Review Board.
Information classified as private, restricted, or confidential by policy or law shall not be entered into an AI system, including as part of an AI query, without approval of the AI Review Board. NOTE: AI Review Board permission is not required to use AI capabilities integrated within approved NDUS CTS (Core Technology Services) or VCSU Enterprise Systems that may contain private, restricted, or confidential data provided all other provisions of this policy are met.
Users shall not attempt to query AI systems to produce information classified as private, restricted, or confidential by policy or law without either a) a VCSU official business need to know, or b) approval of the AI Review Board.
All AI projects, AI queries, and AI systems must comply with relevant data protection laws, SBHE (State Board of Higher Education) policies, including SBHE 1202.3 “Data Privacy;” NDUS (North Dakota University System) procedures, including NDUS 1203.7 “Data Classification and Information Security Standard;” VCSU policies, including V1201-01 “Information Technology Approval Process;” and VCSU IRB (Institutional Review Board) guidelines and processes.

3. Submitting an AI Project for Review

AI projects and AI queries requiring review and approval shall be submitted to the Chief Information Officer. The AI Review Board is responsible for the creation and dissemination of proposal guidelines and forms.

4. Additional Guidance and Training:

VCSU will provide training and resources for employees and students. The VCSU student handbook shall define general guidelines regarding the acceptable use of AI queries as it relates to academic integrity. Faculty may include additional guidance and limitations on applying AI within each course.

5. Complaints, Grievances, or Appeals:

Individuals wishing to report a violation of this policy may contact the Chief Information Officer via email with the subject line “V1203.03 Violation.” The Chief Information Officer shall work with the AI Review Board to respond to the reported violation.
An employee or student who disagrees with the decisions or approval requirements of the AI Review Board concerning a project or query they have submitted for review may request a meeting with the AI Review Board to appeal the decision or approval requirements. The AI Review Board will make a final decision or final approval requirements within 7 business days of the appeal meeting. AI Review Board decisions or approval requirements following an appeal meeting are not appealable.

Co-Sponsored by the VPAA and the CIO

Effective date: March 2024