Many individuals across the university face challenges in using Generative AI tools safely and ethically within the bounds of State law, as well as UW Regent, UW-System, and UWM policies. The goal of this page is to do the following:
- Outline relevant laws and policies and how they apply to Generative AI usage.
- Work to address common questions and issues for students, faculty, staff, and researchers, that apply to a particular policy while presenting UWM-approved responses to typical scenarios involving Generative AI use.
Please note that as privacy and security conditions for Generative AI tools evolve, elements of this page may change.
Applicable Laws and Policies
Chapter 14 of the Wisconsin Administrative Code defines Student Academic Disciplinary Procedures for the UW System, including academic misconduct and integrity violations. These rules apply to unauthorized use of Generative AI tools.
Faculty are strongly encouraged to use CASL’s guidance on Generative AI and include a clear syllabus statement outlining classroom expectations. This helps reduce student uncertainty. Students should review their syllabus and contact their professor with questions. Faculty should avoid using AI detectors due to high false positive rates and consult the Dean of Students Office for any academic misconduct concerns related to Generative AI.
UW System Regent Policy Document 25-5 sets information security expectations for all institutions, covering assets like student records, health data, and confidential information. It establishes a framework for acceptable IT use and protects sensitive data.
Generative AI tools often process all shared data and may be used to train models unless users opt out. Review third-party vendors’ privacy policies before use. Faculty and staff must comply with laws like FERPA; unauthorized sharing of student or sensitive data with AI tools can lead to serious violations and penalties.
The Acceptable Use Policy outlines the expectations of the Board of Regents regarding the acceptable use of IT resources by authorized users and is used to establish the parameters for the use of IT resources.
Generative AI tools on university systems should only be used for approved academic, research, or official purposes. Misuse of Generative AI tools to bypass policies or laws is prohibited. Students must follow syllabus rules, and no one should use AI to create misleading or harmful content. Faculty and staff should not upload confidential university data to public AI platforms.
UWSA Policy 1040 sets standards for handling personal data in the UW System, limiting its collection, use, sharing, and storage to what’s necessary for academic, research, administrative, or legally permitted purposes.
When using Generative AI tools on university systems or networks, users should limit data sharing to use cases that support their academic, research, or official duties. Many tools may use the prompts and responses users provide to train the underlying Large Language Model. Recent cyberattacks have demonstrated the ability to expose a model’s training data. Individuals should always be cautious when sharing information with Generative AI systems. This caution also applies when integrating AI with services like email or OneDrive, which may grant access to all user files, including those shared by others, risking unauthorized data exposure.
UWSA Policy 1031 outlines the various classification levels that apply to data processed, stored, or used by UW System entities, and helps inform which technical, administrative, and physical controls should be applied to protect the data from theft, alteration, loss of integrity, and/or misuse.
Users must not enter confidential or restricted data (e.g., student records, research, financial info) into public AI platforms, which may store or reuse inputs. For example, uploading proprietary research or analyzing student data without safeguards may violate policies like FERPA and lead to serious compliance issues.
UWSA Policy 1039 provides a framework for identifying, assessing, and managing information security risks across the UW System. It aims to protect IT assets by ensuring confidentiality, integrity, and availability, while maintaining compliance with regulations and guidelines.
When using Generative AI, users must assess risks like data leakage, bias, and misinformation. For example, inaccurate AI summaries of sensitive legal documents or unreviewed AI-generated content on university platforms can cause legal or reputational harm. A risk assessment should be completed before integrating AI into teaching, research, or administrative systems.
The UWM Code of Conduct (SAAP 7-3) outlines expectations for integrity, respect, stewardship, and compliance for all individuals acting on behalf of the university.
In practical terms, this policy promotes the responsible use of Generative AI, requiring honesty and transparency in academic and administrative content. Misrepresenting AI-generated work as entirely original could violate principles of integrity and trustworthiness. Uploading non-public university data (such as student records or internal documents) into public AI platforms may breach confidentiality obligations. Additionally, the policy discourages the misuse of university resources, including generative AI tools. No one should use Generative AI tools in ways that compromise the university’s mission or reputation.
The UW System’s Computer Software Ownership Policy outlines who owns software developed within the university.
When using Generative AI, content created independently by faculty or students—like teaching aids or research tools—is typically owned by the creator. However, if developed under a job assignment or university contract, the university may claim ownership. Users should also review licensing and attribution rules, as AI-generated code or content may affect project ownership rights.
The UW System Administrative Policy on Copyrightable Instructional Materials outlines how ownership and usage rights are determined for educational content created within the university system.
When using Generative AI to create instructional materials, ownership typically belongs to the creator if minimal university resources are used. However, if substantial support (like released time, equipment, or staff) is provided, the university may claim ownership or require a formal agreement. Written agreements are encouraged when institutional resources are involved, and transparency is essential in distributing AI-assisted materials.
The Regent Policy Document 4-1 states that “The Board of Regents recognizes the right of individual instructors to set reasonable policies and practices in regard to the use, copying or recording of instructional materials or lectures that are developed as a part of an academic course, event or program.” With a noticeable emphasis on helping to provide access for students with reasonable accommodations. This policy may also apply to the use of AI, including AI notetaking apps.
Additional Resources
Use of AI in Classrooms
Academic Misconduct (Dean of Students) — Suspected misuse of AI in a way that violates instructor expectations may be considered academic misconduct. Procedures for how academic misconduct is handled can be found in the link above.
Resources for Instructors
The Center for Advancing Student Learning (CASL) provides several sample syllabi that can be used to set expectations on AI usage within a course. Additional information on the intersection of teaching and Generative AI can be found in the resources linked below: