Generative Artificial Intelligence Policy

Your Heading

Preamble:

Artificial intelligence (AI) is the use of technology, often implemented through a computer program, which can perform complex tasks, applying logic and reasoning and making decisions which had traditionally only been performed by a human. While this technology has existed for decades, the release of tools such as ChatGPT to the public has increased awareness and wide accessibility of "generative Al" or GAI. GAI has the capabilities of Al, but it can also generate, or create new content (e.g., original text, images, computer code) in response to content it has been trained on and human prompting in everyday language, not computer programming expertise. There are vast applications of GAI in educational settings, as well as uses in medicine, business, human services, technology and many other professions. As such, an understanding of the appropriate use of GAI will be fundamental to academic development. This policy is intended to provide general requirements for the use of Al at UW-Green Bay by faculty, staff and students.

Use of GAI Tools

UW-Green Bay promotes safe, ethical, and effective use of GAI products by students and employees. As the availability and development of GAI tools is rapid, attention should be given to information provided by GBIT updating users of cautions and concerns about products and processes. Employees and students may freely use GAI tools which are integrated into University-provided information technology resources (e.g., SIS, Salesforce). Users of non-University-provided GAI tools must assess that GAI tool to confirm the appropriateness of use and its compliance with the security standards of UW-Green Bay and the Universities of Wisconsin.

Prior to using any AI tool, students and employees must consider the following:

1. All GAI tools used by employees must meet our security and data protection as established in UW System Policy 1031.A: Data Classification. Thus, the use of UWGB-provided tools is recommended. If employing other reputable Al tools, or any other GAI, the user must ensure its security first. That includes, but is not limited to, reviewing the tool's security features, terms of service, and privacy policy. Users should be cautious when using tools developed by individuals or companies without established reputations.
2. Users must apply the same security practices when implementing any GAI tool as they do with all other software applications as required under UW System Policy 1000: Information Security, including our strong passwords requirements, keeping software up-to-date, and following data retention and disposal policies.
3. Users shall not provide access to University GAI tools to those outside of UW-Green Bay without prior approval from the appropriate department and confirmation that third-party use meets security compliance requirements.
4. Unless otherwise confirmed by GBIT, users should not presume that any GAI tool is compliant with FERPA, HIPAA, or data security requirements. Users are under all obligations to protect any information uploaded into the GAI tool and should apply the standards set forth by the Universities of Wisconsin and UW-Green Bay policy. Users shall not upload or share any data that is confidential, proprietary, or protected by regulation without prior approval from the appropriate department.
5. Users who elect to use non-reviewed tools are responsible for any violations of policy resulting from use. Failure to perform appropriate assessment of tools prior to use may result in disciplinary action against the user. 

Additional Information About Student & Employee Use of GAI

No employees or students may be compelled to use GAI tools which require individuals to provide personally identifying information (PII) or other protected data to a third party, whether for a job duty or a class assignment. For reasons of privacy and equity, if GAI is used in courses, tools which are accessible by logging in with UW-Green Bay credentials should be relied upon, but even with those, students also cannot be compelled to input PII and FERPA-protected information.

Because GAI is a constantly evolving technology that may be integrated into existing software or applications, it is not possible to set specific restrictions on all possible uses or applications within the university setting. UW-Green Bay is therefore dependent on individual users to ensure GAI tools are selected and used both ethically and in compliance with other University and System policies. 

Conclusion

GAI tools will continue to advance and become more integrated into all aspects of life. As such, these tools should be accessible, and UW-Green Bay students and employees should be guided to use them productively and ethically. Without proper application of ethical and data protection standards, the use of GAI tools can create issues of academic integrity, data security, and inequity or bias. As such, the above policy should be applied by faculty, staff, and students in any use of GAI within programming at UW-Green Bay. 

Further information on the use of GAI at UW-Green Bay can be found on the AI Policy and Guidelines page. Related documents include:

• RPD 25-3: Acceptable Use of Information Technology Resources
• UW-Green Bay: Security Policy Awareness
• UW-Green Bay: Data Confidentiality Acknowledgement
• UW System Policy 1000: Information Security
• UW System Policy 1031: Data Classification and Protection and Updated Policy (effective 12/1/24) 
• UW System Policy 1031.A: Data Classification
• UW System Policy 1035: IT Asset Management
• UW System Policy 1038: Network Protection
• UW System 1039: Risk Management
• UW System 1042: Threat and Vulnerability Management
• UW System Guidance on Generative AI