UBIT Guidance: Using Generative Artificial Intelligence (AI) Technology at the University at Buffalo

Category: Information Technology

Date Established: January 2,  2024

Responsible Office: Office of the Chief Information Officer

Responsible Executive: Chief Information Officer

The University at Buffalo (UB, university) embraces technological advancement and is committed to providing the campus community with cutting edge technological resources. UB recognizes the potential that Generative Artificial Intelligence (Gen AI) tools have to advance research, instruction, learning and daily office functions at the university. As the University at Buffalo adopts emerging technology, it is important to provide guidelines for the responsible and effective use of these emerging technologies. This document aims to provide an overview of considerations, best practices, and ethical guidelines for using Gen AI tools at the University at Buffalo.

Data Privacy:

When using Gen AI technology individuals should not enter Category 1 or Category 2 data, including non-public research data into Gen AI tools of which the university has not entered into a contract with. Information shared with Gen AI is not private and could expose proprietary or sensitive information to unauthorized parties. Information provided in prompts to Gen AI is often stored by the AI software to be used in future outputs and to train or re-train future Gen AI models.

Gen AI has made it easier for malicious actors to create sophisticated scams, and individuals using AI technology must consider privacy risks, exposure of intellectual property, and the potential threat of data breaches utilizing information provided to Gen AI software. The University at Buffalo works to ensure that all software and applications procured on behalf of the university have the appropriate privacy, security, and accessibility protections in place prior to purchase.

Originality, Plagiarism and Academic Integrity:

AI-generated content can be inaccurate, misleading, or entirely fabricated and may contain copyrighted material.

Bias and Ethical Concerns

There are many ethical concerns regarding the data that is used to train Gen AI models. These include but are not limited to:

• AI Hallucinations – Gen AI tools can "hallucinate" or create text and images that sound and look plausible but deviate from reality or have no basis in fact.
• Bias - Gen AI models often include biases from the data they are trained on. AI systems can exhibit biases that stem from their programming and data sources; for example, machine learning software could be trained on a dataset that underrepresents a particular gender or ethnic group. Many of the biases in Gen AI have yet to be fully identified and understood due to the newness of the technology and emerging capabilities.
• Gen AI software offers extremely limited transparency into the data which is used to train their models.
• Unauthorized inclusion of others works; rights and attributions - In addition to including the works of others without authorization, some Gen AI software will imitate creators without clearly stating so and offering proper credit and attributions, and many artists, authors, programmers, and creators have discovered content generated by AI models that is nearly identical to their own. In many cases, there was never any authorization granted to use their works as part of the Gen AI software’s training data. This can be of particular concern when using AI generated content in research and academic work as it may include another person's work with incorrect or no citation and attribution.

While Gen AI tools can be used to aid in creativity, the originality of the content they produce may be debated. Individuals should be familiar with the terms of use for the software which they are using. Individuals using Gen AI are encouraged to experiment with the software and familiarize themselves with the tool’s capabilities and limitations but should be aware that they are responsible for the content they create when using Gen AI.

• Individuals should always carefully review and edit any information they generate using an AI tool.
• Individuals should not use AI in a way which violates SUNY or university policy, or any existing State or Federal laws.
• Individuals are responsible for ensuring that AI generated content does not infringe on the intellectual property rights of third parties, includes proper citations, and aligns with the university’s policies and procedures.
• Photos created or edited using AI for official UB purposes should follow UB’s Visual Communications Best Practices representing UB’s diversity of race and ethnicity.

When used properly Gen AI tools can be used to aid in accessibility by eliminating unnecessary hurdles in the learning process.

The purpose of this guidance document is to ensure individuals who utilize Gen AI technology are aware of the recommended best practices and associated risks when using such software.

This guidance applies to all University at Buffalo faculty, staff, students, and volunteers who use Gen AI technology for university business or with UB data.

Generative Artificial Intelligence (Gen AI) Tools: Software applications which use machine learning techniques to generate content including but not limited to text, images, and music. AI tools utilize patterns from existing data to produce new content that resembles the input data.

Prompt: Phrases or keywords input into Gen AI software to produce content.

Chief Information Security Officer:

• Conduct periodic security reviews of systems approved for storing and handling protected data.

Generative AI User:

• Develop prompts when using Gen AI which do not use Category 1 or Category 2 data.
• Review and edit the content generated by the AI tool which is being used, to ensure that the information is correct and not plagiarized.

UBIT Software Review Committee:

• Review all incoming software requests for use of Gen AI, and ensure such software meets UB, SUNY, New York State and Federal regulatory requirements.

Vice President and Chief Information Officer
Phone: 716-645-7979
Email: vpcio@buffalo.edu

Information Security Office
Phone: 716-645-6997
Email: sec-office@buffalo.edu

Director of the Office of Academic Integrity
Phone: 716-645-2111
Email: academicintegrity@buffalo.edu

Back to Top