Topic Discovery and Classification for Responsible Generative AI Adaptation in Higher Education

📝 Abstract

As generative artificial intelligence (GenAI) becomes increasingly capable of delivering personalized learning experiences and real-time feedback, a growing number of students are incorporating these tools into their academic workflows. They use GenAI to clarify concepts, solve complex problems, and, in some cases, complete assignments by copying and pasting model-generated contents. While GenAI has the potential to enhance learning experience, it also raises concerns around misinformation, hallucinated outputs, and its potential to undermine critical thinking and problem-solving skills. In response, many universities, colleges, departments, and instructors have begun to develop and adopt policies to guide responsible integration of GenAI into learning environments. However, these policies vary widely across institutions and contexts, and their evolving nature often leaves students uncertain about expectations and best practices. To address this challenge, the authors designed and implemented an automated system for discovering and categorizing AI-related policies found in course syllabi and institutional policy websites. The system combines unsupervised topic modeling techniques to identify key policy themes with large language models (LLMs) to classify the level of GenAI allowance and other requirements in policy texts. The developed application achieved a coherence score of 0.73 for topic discovery. In addition, GPT-4.0-based classification of policy categories achieved precision between 0.92 and 0.97, and recall between 0.85 and 0.97 across eight identified topics. By providing structured and interpretable policy information, this tool promotes the safe, equitable, and pedagogically aligned use of GenAI technologies in education. Furthermore, the system can be integrated into educational technology platforms to help students understand and comply with relevant guidelines.

💡 Analysis

As generative artificial intelligence (GenAI) becomes increasingly capable of delivering personalized learning experiences and real-time feedback, a growing number of students are incorporating these tools into their academic workflows. They use GenAI to clarify concepts, solve complex problems, and, in some cases, complete assignments by copying and pasting model-generated contents. While GenAI has the potential to enhance learning experience, it also raises concerns around misinformation, hallucinated outputs, and its potential to undermine critical thinking and problem-solving skills. In response, many universities, colleges, departments, and instructors have begun to develop and adopt policies to guide responsible integration of GenAI into learning environments. However, these policies vary widely across institutions and contexts, and their evolving nature often leaves students uncertain about expectations and best practices. To address this challenge, the authors designed and implemented an automated system for discovering and categorizing AI-related policies found in course syllabi and institutional policy websites. The system combines unsupervised topic modeling techniques to identify key policy themes with large language models (LLMs) to classify the level of GenAI allowance and other requirements in policy texts. The developed application achieved a coherence score of 0.73 for topic discovery. In addition, GPT-4.0-based classification of policy categories achieved precision between 0.92 and 0.97, and recall between 0.85 and 0.97 across eight identified topics. By providing structured and interpretable policy information, this tool promotes the safe, equitable, and pedagogically aligned use of GenAI technologies in education. Furthermore, the system can be integrated into educational technology platforms to help students understand and comply with relevant guidelines.

📄 Content

Topic Discovery and Classification for Responsible Generative AI Adaptation in Higher Education Diane Myung-kyung Woodbridge1*, Allyson Seba2, Freddie Seba1, Aydin Schwartz1 1*StudyStudio.ai, San Francisco, California, USA. 2Computer Science, University of Southern California, Los Angeles, California, USA. *Corresponding author(s). E-mail(s): diane@studystudio.ai; Contributing authors: aseba@usc.edu; freddie@studystudio.ai; aydin@studystudio.ai; Abstract As generative artificial intelligence (GenAI) becomes increasingly capable of delivering personalized learning experiences and real-time feedback, a growing number of students are incorporating these tools into their academic workflows. They use GenAI to clarify concepts, solve complex problems, and, in some cases, complete assignments by copying and pasting model-generated contents. While GenAI has the potential to enhance learning experience, it also raises concerns around misinformation, hallucinated outputs, and its potential to undermine critical thinking and problem-solving skills. In response, many universities, colleges, departments, and instructors have begun to develop and adopt policies to guide responsible integration of GenAI into learning environments. However, these policies vary widely across institutions and contexts, and their evolving nature often leaves students uncertain about expectations and best practices. To address this challenge, the authors designed and implemented an automated system for discovering and categorizing AI-related policies found in course syllabi and institutional policy websites. The system combines unsupervised topic modeling techniques to identify key policy themes with large language models (LLMs) to classify the level of GenAI allowance and other requirements in policy texts. The developed application achieved a coherence score of 0.73 for topic discovery. In addition, GPT-4.0-based classification of policy categories achieved precision between 0.92 and 0.97, and recall between 0.85 and 0.97 across eight identified topics. 1 arXiv:2512.16036v1 [cs.AI] 17 Dec 2025 By providing structured and interpretable policy information, this tool promotes the safe, equitable, and pedagogically aligned use of GenAI technologies in edu- cation. Furthermore, the system can be integrated into educational technology platforms to help students understand and comply with relevant guidelines. Keywords: Generative AI in Education, Generative AI Policies in Education, Ethics of Generative AI in Education, Tools for Moderating and Configuring Generative AI in Education, Educational Technologies (EdTech) 1 Introduction Since the release of OpenAI’s ChatGPT [1], a generative artificial intelligence (GenAI) using a large language model (LLM), in November 2022, various companies, including Cohere [2], Anthropic [3], Google [4], and Meta [5], have introduced powerful LLM models and their updates. These models, and the applications built on them, are trans- forming many aspects of daily life and work. These models, and the applications built on them, are transforming many aspects of daily life and work—from customer service chatbots and software development to marketing with AI-generated media and even drug discovery [6]. The use and applications of LLMs have largely impacted educa- tion and learning as well. Thanks to their ability to generate personalized responses, deliver real-time feedback, and enhance efficiency, students have rapidly adopted tools powered by these models [7]. Despite the benefits of GenAI in education, there have been many concerns includ- ing academic misconduct, hallucination, biases, legal and ethical use, privacy, and security [8]. Crucially, without clear guidance and best practices, GenAI tools may hinder rather than help learning—potentially weakening students’ problem-solving abilities, critical thinking, and foundational understanding [9]. Recent research shows that dependency on GenAI tools can cause memory loss, procrastination, and worsen academic performance [10]. In education, higher education institutes first started discussing incorporating GenAI policy and guidelines into student honor codes and course syllabi. As GenAI is new to many, higher education institutes organized committees including experts in artificial intelligence, education, policy, ethics, and other disciplines to endeavor to provide templates, examples, and general guidelines [11]. As AI policies are established based on ethical considerations, learning objec- tives, and pedagogical approaches of the discipline, instructor, course, and even an assignment level, they vary at the institution, department, course, lesson plan, and assignment levels [12][13]. This variability can result in vague or inconsistent guidance, contributing to confusion and inequity in student learning experiences. As such, clear articulation of expectations including allowed usage, attribution, data validation, and cautions on information release is critical

This content is AI-processed based on ArXiv data.