601.15 - Ethical and Effective Use of Artificial Intelligence (AI)

1. Purpose

This policy establishes a unified framework for the ethical, effective, and mission‑aligned use of artificial intelligence (AI) at the University of Arkansas Fort Smith (UAFS). It ensures that AI tools are used in ways that advance the University’s mission, protect the integrity of teaching and research, and support positive student outcomes.

This policy is grounded in UA System Policy 285.2 (Artificial Intelligence – Responsible Use) and UA System Policy 285.3 (Artificial Intelligence and Automated Decision Tools), both of which implement Act 848 of 2025 (Ark. Code Ann. § 25-1-128).

2. Scope

This policy applies to all students, faculty, staff, and affiliates of UAFS; all academic, research, and administrative units; and all AI tools used for University purposes, including both institutionally provided and externally accessed tools.
Departments and units may adopt more restrictive guidelines consistent with this policy and UASP 285.2 and 285.3.

3. Guiding Principles

All AI use at UAFS must adhere to the five systemwide principles established in UASP 285.2 and the four institutional principles below.

Integrity and Transparency: AI must be used in a manner that upholds academic integrity, professional ethics, and transparency of purpose.
Human Oversight: AI must augment, not replace, human judgment and accountability.
Non-Discrimination: AI systems must be used in ways that promote fairness and prevent bias or discrimination.
Data Stewardship: Use of AI must protect privacy, confidentiality, and intellectual property.
Compliance and Accountability: All AI activities must comply with applicable laws, accreditation standards, and Board of Trustees and UA System policies.

4. Definitions

Artificial Intelligence (AI): Technologies that perform tasks requiring human-like cognition, including generative AI, machine learning, chatbots, image generators, and predictive analytics.
Generative AI: Tools that create text, images, audio, code, or other content based on user prompts.
AI-Assisted Work: Academic, research, or administrative work in which AI contributes to idea generation, drafting, analysis, or decision support.
Automated Decision Tool: A system or service using artificial intelligence that has been specifically developed or modified to make, or be a controlling factor in, consequential decisions, as defined in UASP 285.3.
AI Oversight Committee (AIOC): The institutional body designated under UASP 285.2, Section V.C, responsible for reviewing emerging AI technologies, maintaining the approved tools inventory, and ensuring ongoing policy compliance.
High-Risk AI: An AI tool used in decisions affecting student academic standing, employment, financial aid, health, or civil rights, or that has the potential to significantly affect institutional operations or individual rights. Such tools require AIOC review and documented risk mitigation prior to institutional deployment.
Sensitive Data: Data classified as Highly Sensitive or Internal under UAFS Policy 902.2 (Data Classification Policy and Procedure).

5. Acceptable Use of AI

AI may be used when:

it supports student learning and academic achievement (when it complies with course, program, and institutional rules);
enhances scholarly productivity;
improves administrative efficiency without compromising fundamental fairness;
respects privacy and data security requirements;
is transparent to affected users; does not replace essential human judgment;
is accessible to all students, including those with limited prior exposure to AI.

Examples of acceptable use include:

brainstorming, outlining, or revising academic work (within parameters established by the relevant faculty);
literature scanning, coding assistance, or data analysis (see copyright language in Section 11);
drafting administrative communications;
accessibility support such as captioning and text simplification;
research assistance using approved tools;
and AI-enhanced tutoring or writing support provided through University services

6. Prohibited Use of AI

AI may not be used to:

submit AI-generated work as one's own when prohibited by course or assignment guidelines;
enter or upload sensitive or confidential data into unapproved AI tools;
automate grading or academic evaluation without active human oversight;
make final decisions in advising, admissions, financial aid, conduct, or employment;
generate deceptive, harmful, or discriminatory content;
circumvent learning outcomes or misrepresent authorship;
fabricate research data, citations, or scholarly findings;
violate IRB protocols, export control regulations, or data management requirements;
replace human-centered advising, mentoring, or student support relationships;
express a personal political opinion to an elected official or engage in lobbying, unless within the scope of regular job duties, consistent with UASP 285.3, Section V.

7. Academic Use and Course-Level Expectations

7.1 Faculty Authority

Faculty determine the extent to which student AI usage is permitted in their courses, consistent with UASP 285.2, Section III.A. Expectations must be communicated clearly in course syllabi. The AIOC will provide sample syllabus language (see Appendix A).

7.2 Student Responsibilities

Students must follow course-specific AI guidelines, disclose AI use when required, and understand that unauthorized AI use may constitute academic misconduct under existing University policies.

7.3 Support for UAFS Students

The University is committed to providing students with the resources, instruction, and support needed to develop AI literacy, understand academic integrity expectations in the context of AI use, and engage with approved tools effectively and responsibly.

The specific forms of support will be determined by the AIOC in coordination with relevant academic and student affairs offices. They will be updated as institutional needs and available tools evolve. More detailed student support programming will be addressed pursuant to Section 18 of this policy.

8. Research and Scholarly Use

AI may support research activities including data analysis, coding and computational modeling, literature review and synthesis, and drafting or revising scholarly text.

Researchers must:

maintain responsibility for the accuracy and integrity of their work;
disclose AI use when required by journals, funders, or disciplinary norms;
avoid using AI to fabricate data, citations, or results; follow data security requirements when processing research data;
comply with IRB protocols, export control regulations, and data management plans;
use only AI tools approved for research involving sensitive or regulated data, consistent with UASP 285.2, Section III.B.

9. Administrative and Operational Use

AI may support administrative tasks such as drafting communications, scheduling and workflow support, data analysis and reporting, and decision support tasks.

Human review is required for all AI-assisted recommendations.
Units must:

use only institutionally approved tools;
comply with UASP 285.3 regarding automated decision tools;
AI-supported advising or predictive analytics must be evaluated for disparate impact on any significant group of UAFS students or employees, consistent with UASP 285.2, Section III.C.

10. Data Privacy and Security

All AI use must comply with UAFS Policy 902.1 (Data Governance Policy and Procedure) and the Data Stewardship principle of UASP 285.2.

Users must not input data classified as Highly Sensitive or Internal under UAFS Policy 902.2 (Data Classification Policy and Procedure) into unapproved AI tools, including:

FERPA-protected student records
HIPAA-protected health information
confidential personnel data
payment card data
export-controlled or IRB-regulated research data

Users are responsible for knowing the classification of data they handle before using AI tools to process, analyze, or transmit it.

AI-generated output should be treated as informational and not authoritative unless verified by a qualified individual.

The AIOC will maintain an approved list of AI tools verified to meet the security and privacy requirements of UAFS Policy 902.1 (see Section 14.2).

11. Intellectual Property and Copyright

UASP 285.2, Section IV.D, delegates intellectual property questions arising from AI use to individual institutions. The following provisions constitute UAFS's institutional response to that delegation. The AIOC shall monitor developments in copyright law and recommend updates as warranted.

11.1 Input of Third-Party Copyrighted Materials

Inputting third-party copyrighted materials into AI tools may implicate the exclusive rights of copyright holders, including reproduction and the creation of derivative works. Whether such use constitutes infringement or qualifies as fair use is fact-specific and, in the context of AI tools, legally unsettled.

As a precautionary measure, users should not input third-party copyrighted materials into AI tools without authorization from the rights holder or documented fair use analysis.3 Users with questions should consult the UAFS Library.

11.2 Ownership and Copyright Protection of AI-Generated Outputs

Content generated solely by an AI system, without meaningful human authorship, is not protected by copyright under United States law. AI-assisted works in which a human makes meaningful, identifiable creative choices may be protected, but only to the extent of the human-authored elements. Users must not assert copyright ownership over purely AI-generated content.

11.3 Institutional Work Product

UA System Board Policy 210.1 governs intellectual property ownership for work created by University employees. Under that policy, faculty generally retain copyright in scholarly works, including textbooks, journal articles, and creative works, produced in the course of normal teaching and scholarly activity.

The University retains ownership of commissioned works, research data, and computer software as defined in UABP 210.1, including those produced with AI assistance. AI-generated content that lacks sufficient human authorship to qualify for copyright protection, as established in Section 11.B of this policy, is not a copyrightable Work under UABP 210.1.

11.4 Research, Scholarship, and Publication

AI may not be used to fabricate data, citations, or results. For research conducted under federal funding, this constitutes research misconduct subject to applicable agency regulations.

Faculty, staff, and students must comply with the authorship and AI disclosure requirements of the journals, publishers, funding agencies, and disciplinary bodies to which they submit work.

Requirements vary; users are responsible for knowing the standards applicable to their specific submissions.

11.5 Vendor Intellectual Property Terms

Before any AI tool is approved for institutional use, the Office of Information Technology shall review the vendor’s contractual terms, including provisions governing ownership of user inputs and outputs, data use for training purposes, confidentiality, and data security.

Institutionally contracted enterprise agreements are preferred over free consumer version of AI tools, as enterprise terms typically provide stronger protections for institutional and user data. This review is conducted pursuant to UAFS’s standard vendor contracting procedures.

12. Accessibility and Accommodations

AI tools must meet applicable accessibility standards, including the Americans with Disabilities Act and Section 508 of the Rehabilitation Act, and must not replace required human accommodations.

The AIOC shall evaluate proposed tools for potential barriers to students with disabilities as part of the approval process. Students may use AI as part of an approved accommodation when authorized by Disability Services, but such tools should be functionally limited to the specific accommodation intended and not available for broader use that may allow for assignments to be inappropriately completed by the AI tool.

The Student Disability Service office (SDS) shall work with faculty to ensure that the SDS Office retains its autonomy to issue authoritative accommodation letters while also tailoring any accommodation involving AI narrowly to protect the academic integrity of the classroom.

13. Transparency and Disclosure

Users must disclose AI involvement when required by course policy, research norms, or unit guidelines. Administrative units must inform students when AI is used in decision-support processes that affect them. Plain-language versions of AI policies will be published to ensure accessibility.

14. Governance and Oversight

14.1 AI Oversight Committee (AIOC): Composition and Authority

The AIOC is established pursuant to UASP 285.2, Section V.C. It oversees implementation of this policy and reports annually in writing to the Chancellor. Members are appointed by the Provost, in consultation with the Chief Information Officer and the VCFA. The AIOC meets at minimum quarterly.

Voting Members (10):

The AIOC will be comprised of no more than ten persons, representing faculty, staff and students. The following areas of representation are to be considered (One person may represent more than one area):

Faculty Senate
Student Government Association
Staff Senate
Center for Teaching and Learning (CTL)
Information Technology (IT) and/or Instructional Support (IS)
Human Resources
Library
Student Success/Enrollment Management (Staff Representative)
Dean's Council and/or Academic Leadership
AI Education Subcommittee (Faculty)
Advancement
Office of Institutional Effectiveness

Voting members serve two-year staggered terms. The student representative serves a one-year term. The AIOC Chair may appoint ad hoc members with relevant expertise as needed or invite advisors with specialized knowledge on particular topics.

14.2 Approved Tools Process

The Office of Information Technology leads the evaluation of AI tools for institutional use, assessing each against standards for data privacy, cybersecurity, accessibility, vendor intellectual property terms, and potential for bias or disparate impact.

Based on IT's recommendation, the AIOC grants or denies approval and maintains final authority over the approved tools list, which IT publishes and maintains. IT shall escalate material changes in vendor terms or security posture to the AIOC for review.

Before purchasing or contracting for any AI tool, units must work with IT and the AIOC to receive approval before deploying.

Persons using non-approved AI tools on university equipment or on personal devices when connected to the university network assume the risks inherent therein and are subject to all of the principles and guidelines expressed in this policy and relevant state and system policies.

14.3 Incident Reporting

Units or individuals who become aware of an AI-related incident, including discriminatory output, significant errors affecting students or employees, or unauthorized data exposure, shall report the incident to the Director of Information Technology Services and the AIOC Chair within two (2) business days.

Incidents involving data exposure or a suspected security breach must also be reported immediately in accordance with UAFS Policy 903.9 (Cybersecurity Breach/Incident Response). The AIOC shall determine appropriate remediation and coordinate with relevant University offices.

15. Training and Support

The University will provide training and support resources to help faculty, staff, and students use AI tools effectively, ethically, and in compliance with this policy.

The specific forms of training and support will be determined by the AIOC in coordination with relevant academic and administrative offices; the offerings will be updated as institutional needs and available tools evolve.

The AI Education Subcommittee of the AIOC will play a significant role in coordinating trainings for campus populations.

16. Policy Compliance

Violations of this policy may result in academic integrity proceedings (students), student conduct proceedings (students), corrective or disciplinary action (employees), or removal of access to AI tools.

The University will distinguish between intentional misuse and novice errors arising from lack of AI literacy and will offer educational interventions for the latter when appropriate.

Use of AI in research must comply with Institutional Review Board (IRB) protocols, export control regulations, data management requirements, and institutional policies where applicable.

17. Review Cycle

This policy will be reviewed annually by the AIOC and updated as needed to reflect technological, legal, and institutional developments, consistent with
UASP 285.2, Section VII.A.

18. Supplemental Guidelines

This policy establishes the institutional framework for AI use at UAFS. The AIOC may develop and adopt supplemental guidelines for specific functional areas as needed.

Supplemental guidelines must be consistent with this policy and are subject to approval by the relevant institutional authority.

Until supplemental guidelines are adopted, this policy governs all campus AI use. AI tools identified as High-Risk under Section 3 require written AIOC review, documented risk mitigation, and Cabinet approval prior to purchase, implementation, or use.

19. Other Applicable Policies

Appendix A: Sample Syllabus Language

This language will be developed by the AIOC and Faculty Senate by end of Fall 2026. It will include three tiers: Unrestricted Use, Mixed Use, and Total Prohibition with approved language for each.

Appendix B: Course-Level AI Use Framework

The following framework, adapted for UAFS from models in use at peer institutions, is provided as guidance for the Faculty Senate's development of detailed course-level guidelines. It does not constitute policy.

Faculty/staff expectations regarding AI use must be included in the course syllabus and may be reinforced through course materials, assignment instructions, research guidelines, the learning management system (LMS), or other standard communications.

Types of permitted use to be specified in the syllabus:

Unrestricted Use: Students may use AI technology for any learning, creation, or analysis without restriction.
Mixed Use: Students may use AI for some purposes, with specifics provided clearly and in a timely manner.
Total Prohibition: Students may not use AI technology for any learning, creation, or analysis.

Faculty should consider the following use cases when describing any limitations:

Learning: AI providing definitions, facts, or summaries;
Creating: AI generating text, images, or other content;
Analyzing: AI examining patterns or data;
Controlling: AI operating other software or systems.

Appendix C: Consulted Policies and Resources

Frostburg State University AI Policy
Albany State University AI Policy
Idaho State University AI Policy
Utah State University AI Policy
University of Tennessee – Martin AI Policy
University of Maryland – Baltimore AI Policy
Fort Lewis College AI Policy
University of Nebraska – Kearney AI Policy
U.S. Copyright Office, Copyright and Artificial Intelligence Reports (Parts 1-3, 2024-2025)
ICMJE Recommendations on AI Use by Authors (2025)
UA System Policy 285.2 (April 7, 2026)
UA System Policy 285.3 (December 4, 2025)