In just a few short years, generative artificial intelligence (Gen AI) has sparked an upheaval in higher education, disrupting long-standing practices and prompting wide-spread reflection on how teaching and learning should evolve in response. Since the public release of large language models in late 2022, faculty and students have experimented, resisted, embraced, and debated their place in teaching and learning. Recent national surveys show that most students now use Gen AI tools in some capacity, as shown in the 2025 AI in Education Trends Report by Copyleaks (2025), which reports that nearly 90% of student respondents confirm their use of AI tools for academic purposes. This included using Gen AI for tasks like brainstorming, drafting, and summarizing, and 73% of student respondents reported using Gen AI tools more this year than they did the year before. A recent Inside Higher Ed and Generation Lab survey stated that 87% of student respondents reporting they know when to use Gen AI tools and 41% reporting that they knew when to use Gen AI because of guidance provided by their instructors through the course syllabi, with an additional 35% knowing when to use Gen AI because it was discussed in class (Mowreader, 2025). Results from a February 2025 survey administered by the Higher Education Policy Institute and Kortext also indicate a high number of students using Gen AI tools, including 92% of student respondents saying they are using Gen AI tools in general and 88% reporting the use of Gen AI tools for academic assessments specifically (Freeman, 2025).
Faculty adoption of Gen AI has been on the rise since 2022 but has not been as high as student adoption. Faculty responses to Gen AI vary widely, from outright prohibition to intentional integration (Mulford, 2025). A 2024 survey conducted by the American Association of Colleges and Universities and Elon University’s Imagining the Digital Future Center of 337 higher education leaders reported that less than 40% of faculty were using Gen AI tools as part of their roles (Rainie & Watson, 2025). These same leaders report “faculty unfamiliarity with or resistance to GenAI” tools as one of the most often noted barriers to Gen AI use across their campuses (Rainie & Watson, 2025, p. 3). Similarly, a 2025 survey conducted by the Digital Education Council found that 61% of faculty reported using Gen AI in their teaching, with 88% saying they do so minimally (Grovergrys & Rettler-Pagel, 2025). Faculty have reported mixed attitudes towards the integration of Gen AI tools into teaching and learning, expressing both concerns related to cheating and hindering the learning experiences, as well as positive outcomes that Gen AI tools may bring to education, like opportunities for personalized learning and individualized support for students (Mulford, 2025).
Amid this rapid change, higher education institutions are recognizing the importance of offering clear, consistent guidance to help campus communities use Gen AI safely, ethically, and effectively. However, many campuses are still in the process of crafting policy language for their institutions. In a recent Inside Higher Ed survey of campus provosts, half of all respondents said, “their institutions are actively developing policies for AI use but have not fully implemented those policies” and only 14% report having a comprehensive AI policy in place (Whitford, 2025, p. 12). There is also tension in many policies between enthusiasm and concern, specifically related to the use of Gen AI tools to encourage creativity, increase personalization of student learning experiences, and increase instructor efficiency. At the same time, institutions must remain aware of the risks these tools bring, including academic integrity violations, biased output, misinformation, and the possibility that students may utilize these tools to evade learning rather than enhance it. Institutions of higher education have reacted in varied ways, including university wide policies that include honor codes, academic integrity updates, and Gen AI guiding principles documents (Rainie & Watson, 2025; Syracuse University, 2025; University of Notre Dame, 2023), as well as course-level policies that typically reside within syllabi documents (Cornell University, n.d.). Some of these policies provide language faculty must adopt, may choose to adopt, or guidance on language they should then craft on their own. The best solution is still emerging, as the speed of advancement with Gen AI technology makes it challenging for even leading experts to develop policy or guidance that stays current.
This article examines one University’s experience developing a consistent yet flexible Gen AI syllabi policy that establishes clear, campus-wide expectations while allowing course-level customization to reflect disciplinary differences and faculty comfort with Gen AI use. It also highlights how the institution tracked and refined the policy’s implementation over the past two and a half years to keep pace with the rapidly evolving technological landscape.
One University’s Gen AI Syllabi Policy Journey
A small private University’s work over the past two and a half years to design and implement a campus-wide Gen AI syllabi policy offers a meaningful look at how institutions can balance structure with flexibility while also monitoring patterns of Gen AI use across courses. The approach taken by this institution blends clear, institution-level expectations with room for faculty to tailor guidance to the needs of their disciplines and assignments. Using course syllabi as the primary vehicle for this communication proved especially effective, as these documents are both widely accessible and serve as a direct, trusted channel for conveying expectations to students. This model also differs from the approach taken at many institutions by placing both the university-level policy and the course-specific guidance in one coherent, easy-to-access location.
In 2023, an interdisciplinary working group of faculty, staff, librarians, academic leaders, and students was organized and asked to create guidance on Gen AI use for the University. The resulting work included recommendations for minor updates to the existing academic integrity policy, the creation of an ongoing cross-campus AI Advisory Board, the creation of an AI Toolkit to act as a resource on proper Gen AI use across campus roles, as well as initial language to be included in all course syllabi related to the use of Gen AI (Appendix A). These recommendations were approved and implemented for the 2023/2024 academic year. The initial syllabi language ensured transparency and consistency for students. It also aimed to balance academic integrity concerns with safe and ethical Gen AI use, while honoring faculty autonomy and disciplinary differences. Faculty were asked to keep the shared language unchanged, but they could select one of three statements that clearly defined how Gen AI would be handled in their courses. The options specified Gen AI use as prohibited, meaning it was not allowed under any circumstance; permitted, allowing students to choose whether to use Gen AI; or required, where students were expected to use Gen AI and given support to do so.
Over time as perceptions, awareness, and comfort with Gen AI evolved, the Gen AI syllabi policy language evolved as well in both language and intent. After widespread professional development on the use of Gen AI within educational settings, and as Gen AI tools became more embedded in professional practice and scholarly workflows, the conversation on campus expanded to include topics on how Gen AI might enrich learning, encourage creativity, and promote higher-order thinking. It also became clear that the existing policy did not reflect the nuance of Gen AI adoption. One broad Gen AI statement per course was not adequate, but instead many faculty wanted to see a more detailed description of how Gen AI might be used differently within the same course for different activities and assignments. It also became clear that the ability to fully prohibit student use of Gen AI tools was not possible and was not in line with institutional values. Gen AI came to be viewed as a tool, one that could be used to both help and hinder learning, much like many other tools available to students already. Therefore, the Gen AI syllabi language was revised by the AI Advisory Board in the summer of 2024 and the second iteration was implemented in the 2024/2025 academic year (Appendix B). The revisions included a declarative University policy that all students were permitted to use Gen AI tools in certain ways, including tasks like personal study, to further explore course topics, to brainstorm ideas, and to seek assistance from campus services that would enhance their learning experiences across all their courses. It also required faculty to describe the ways Gen AI might be required, allowed, and/or prohibited in their courses at the assignment level rather than for the entire course overall.
In the summer of 2025, the Gen AI syllabi policy language was again amendment, but this time with a minor addition to the general University policy to add additional context and justification for the University-wide Gen AI policy statement, as it became clear that both faculty and students wanted to understand the deeper motivation behind the University’s stance on Gen AI use (Appendix C). The additional text added is provided below and this updated version of the policy went into effect for the 2025/2026 academic year:
The University believes it is important to prepare students for a future in which artificial intelligence (AI), including generative AI, is embedded in nearly every profession. As such, the University encourages the thoughtful use of AI tools to support learning, foster curiosity, and develop the skills needed to thrive today and in the future. Equally important is helping students develop the human judgment to know when and why to use AI, along with the ethical awareness, critical thinking, and communication skills necessary to apply these tools responsibly and effectively.
These versions of Gen AI syllabi policy language have now been used for the past two and a half years across the entire University. The implementation of the policy has been tracked over that same period, which provides insight into the adoption of Gen AI tools at the course level across the campus.
Over the period of five consecutive semesters, syllabi were collected from across the University. This task was part of a standard protocol where all syllabi are submitted by faculty to the respective dean or departmental office each semester. Although some syllabi are occasionally missing from these submissions, the collection represents a large and essentially random sample of the syllabi in use, providing a strong foundation for further analysis. Each syllabus was coded by multiple coders who were provided detailed instructions on how to compare the language present in each course syllabus with the policy language provided by the campus in that given semester. There was very little judgement needed on behalf of the coder, as it was a review if the language was present or not. This data was then analyzed at the aggregate level across the University. This data analysis approach yields a unique dataset showing real faculty behavior, not just stated intent, in a large course sample of over 3,300 course syllabi.
Table 1 shows the overall number of sections where syllabi were collected, number of syllabi with no Gen AI policy included, and the percentage of courses with each type of policy out of the total syllabi collected. This data was then broken down by which of the three categories the policy included, either a prohibited stance on Gen AI use, permitted/allowed use of Gen AI, or a required use of Gen AI. Faculty teaching courses during the fall 2023 or spring 2024 semester were only able to select one of these options, while faculty teaching during the fall 2024, spring 2025, or fall 2025 semesters were able to designate the level of Gen AI use at the assignment level within their courses. Therefore, the percentage for these semesters is greater than 100% total.
Table 2 presents a closer look at how Gen AI prohibition was applied across the fall 2024, spring 2025, and fall 2025 semesters, offering a more nuanced picture than a simple count of courses labeled “AI-prohibited.” Because faculty could choose multiple categories to reflect different expectations across their assignments during these semesters, the data reveals important distinctions within the group of courses that included any form of prohibition. Some faculty used the prohibited category in a targeted way, applying restrictions only to particular assignments while still allowing or even requiring Gen AI use in other parts of the course. Other faculty selected only the prohibited category, signaling that Gen AI use was not permitted in any context of the course. This layered view demonstrates that “AI-prohibited” did not always mean a complete ban, and it highlights how instructors were beginning to adopt more assignment-specific and intentional approaches to Gen AI integration during this period.
Data from the University’s rollout of its campus-wide Gen AI syllabi policy shows a steady rise each year in the percentage of courses identifying as Gen AI-permitted/allowed and Gen AI-required, while the number of courses choosing some amount of prohibition has remained relatively stable. The increase in courses in the allowed/permitted category starting in the fall 2024 semester may be a result of faculty being able to indicate more than one category in their policy during that time.
The fall 2025 semester marked an especially notable shift, with a significant increase in courses designating Gen AI use as a required component of learning. This change suggests growing confidence among faculty in integrating Gen AI directly into coursework rather than treating it as an optional or peripheral tool. At the same time, the proportion of courses where Gen AI was entirely prohibited dropped sharply. Fewer than one quarter of fall 2025 courses fell into this category, a major shift from fall 2023 and spring 2024, when more than half of all courses prohibited Gen AI outright. This movement away from blanket restriction and toward more nuanced or encouraged use reflects a broader cultural change on campus. The University’s practice of revisiting the policy each year and making thoughtful adjustments as technologies and expectations shifted has been key to sustaining clarity, relevance, and trust among faculty, staff, and students.
Lessons Learned and Recommendations
The lessons from this experience reach well beyond the boundaries of a single institution and speak to a broader shift unfolding across higher education. As Gen AI capabilities continue to grow and reshape the academic landscape, institutions are challenged to design policies that provide steady guidance while remaining open to change. When expectations are clearly articulated at the university level, students benefit from a coherent learning environment rather than being confronted with a confusing mix of rules that vary from one class to the next. Such clarity supports fairness, reduces unnecessary anxiety, and helps students build healthy habits around technology use.
At the same time, these policies cannot remain static. New tools appear rapidly, disciplinary needs vary, and faculty continue to refine their understanding of how Gen AI can strengthen or complicate learning. A policy that is too rigid risks becoming obsolete, while one that remains responsive encourages thoughtful experimentation and supports faculty in making informed decisions that suit the goals of their courses. Tracking how these policies are adopted across the curriculum offers a useful snapshot of emerging patterns, highlighting where Gen AI is appearing in learning activities, where it is restricted, and how quickly practices are changing.
Still, numbers alone do not tell the full story. The number of courses allowing or prohibiting Gen AI use cannot capture the pedagogical reasoning behind a faculty member’s decision, nor should the absence of Gen AI in a course be interpreted as a sign of resistance or a lack of knowledge on the part of the faculty member themselves. Many instructors make intentional choices, rooted in learning outcomes, disciplinary norms, or concerns about skill development, that lead them to limit or prohibit Gen AI for certain assignments. Understanding Gen AI adoption in higher education therefore requires attention to both quantitative indicators and the rich, context-specific judgment that faculty bring to their teaching. Future research in this area should pair both the quantitative analysis of Gen AI syllabi policy adoption with qualitative methods to understand the reasons more deeply behind these policy selections. In addition, student feedback on their experiences in courses with varying Gen AI policies would further this research and enhance the conversation around Gen AI adoption within higher education.
In the end, effective Gen AI syllabi policies are not about control, but about cultivating a shared vision that empowers faculty and students to explore these technologies thoughtfully, ethically, and in ways that enhance learning rather than undermine it.
Author Note
Katie M. Sabourin - https://orcid.org/0000-0003-3785-2955
Katie M. Sabourin is the assistant vice president for digital learning in the DePeters Family Center for Innovation and Teaching Excellence at St. John Fisher University, Rochester, NY.
Conflicts of interest
There is no known conflict of interest to disclose.
Correspondence
Correspondence concerning this article should be addressed to Katie Sabourin, DePeters Family Center for Innovation and Teaching Excellence, St. John Fisher University, 3690 East Ave. Rochester, NY 14618. Email: ksabourin@sjf.edu