Universities are increasingly embracing generative AI in higher education, but many policies lack comprehensive guidance on privacy, ethics, and STEM applications while placing significant burdens on faculty to revise pedagogical practices.

Objective: The main goal of this study was to analyze institutional policies and guidelines regarding generative artificial intelligence (GenAI) across US higher education institutions to understand how they address the use of GenAI in teaching and learning environments.

Methods: The researchers analyzed documents from 116 US universities classified as high research activity (R1) institutions regarding their GenAI policies. They collected publicly available policy documents and guidelines from university websites between October and November 2023, resulting in 141 documents. The research team used qualitative coding methods to identify themes and patterns across these institutional policies, developing a codebook with five main areas of analysis and various subcodes. They assessed the extent to which universities encouraged or discouraged GenAI use and what specific guidance they provided to faculty.

Key Findings:

• The majority of universities (63%) encourage the use of GenAI, with 41% providing detailed guidance for classroom integration.
• Over half of the institutions (56%) provided sample syllabi statements and 50% offered sample GenAI curriculum activities to help instructors incorporate GenAI in teaching.
• More than half of institutions (54%) provided guidance on designing assignments that discourage GenAI use, showing a balanced approach to integration.
• Only half of the institutions (50%) addressed GenAI use in STEM fields, with most guidance focusing on writing activities rather than coding or other technical applications.
• More than half of institutions (52%) discussed ethics of GenAI, including Diversity, Equity, and Inclusion (DEI) considerations.
• Privacy concerns were addressed by 60% of institutions, but only 18% specifically mentioned legal implications like FERPA violations.
• One-third of institutions (30%) suggested using GenAI for lesson planning, while 44% discouraged the use of GenAI detection tools due to their unreliability.
• Many policies embraced "flip classroom" approaches that fundamentally change teaching methods to accommodate GenAI.

Implications: This research reveals a significant shift in higher education's approach to GenAI—moving from initial resistance to widespread acceptance and integration. The findings suggest that institutions are rapidly developing frameworks for GenAI use that may substantially transform teaching practices. The study highlights concerns that these policies often place significant burdens on faculty to redesign courses and assessment methods without sufficient consideration of long-term pedagogical implications, privacy concerns, or intellectual property issues.

Limitations: The study was limited to R1 institutions in the United States and only analyzed publicly available documents. The data collection period (October-November 2023) means more recent policy changes were not captured. The research focused on stated policies rather than actual implementation or effectiveness, and did not include perspectives from faculty or students about these policies. Additionally, the study did not fully explore the long-term pedagogical implications of the recommended changes.

Future Directions: The authors suggest that future research should:

• Explore GenAI policies in higher education contexts outside the US
• Investigate instructors' awareness, perceptions, and adoption of these policies
• Assess the long-term impact of GenAI on teaching, learning, and intellectual development
• Evaluate whether the embraced practices produce desired learning outcomes
• Examine GenAI use specifically in STEM fields, which received limited attention in current policies
• Study the ethical and privacy implications of GenAI use more comprehensively
• Analyze how GenAI impacts research practices, which was rarely addressed in current policies

Title and Authors: "Generative artificial intelligence in higher education: Evidence from an analysis of institutional policies and guidelines" by Nora McDonald, Aditya Johri, Areej Ali, and Aayushi Hingle Collier.

Published On: January 17, 2025

Published By: Computers in Human Behavior: Artificial Humans