OPINION article
Front. Educ.
Sec. Digital Learning Innovations
Abstract
It is no secret that artificial Intelligence (AI) based tools have transformed and are still transforming undergraduate education for medical and allied health professionals. AI tools assist in a wide range of tasks associated with learning. For instance, tutor chatbots, AI platforms for diagnosis, science writing, and AI assisted learning skills such as intubation or surgical skills training (1). In physiology education, students can use AI generative language models for tasks such as creating study schedules, studying challenging concepts like Wigger's diagram, and organizing or completing student notes. Students can even ask these large language models (LLMs) to create multiple choice questions on a specific topic to test their knowledge. Furthermore, students can use them to identify various study resources such as YouTube videos (2). All positive learning aspects. As such, students have quickly adopted various AI tools to foster their learning. For instance, students with disabilities have benefited from using AI tools to facilitate learning. Students with dyslexia have used AI tools that convert text to speech, which has resulted in reduced reading time and enhanced comprehension (3). Research has shown that overall students perceive AI tools conducive to learning and have a favorable attitude towards them. Surprisingly, students have also reported that AI tools are accurate and credible (4). Students use these tools for efficient and effective learning (5), but the rapid acceptance of AI tools from students creates an impetus for professors to thoughtfully integrate AI tools in their pedagogical practices, such as project-based learning or Socratic methods. In this opinion piece, we will discuss the importance of including an AI-usage framework in course syllabi that promotes AI literacy and still achieves student learning outcomes. A secondary goal of this article is to consider a counterpoint that a campuswide AI framework is necessary to deter plagiarism and seamless integration of AI in curriculum. Since physiology courses are housed in undergraduate universities (basic science departments) and institutes offering medical degrees or allied health professional degrees such as nursing or dentistry, we have included medical education while discussing campus-wide frameworks. While AI tools bring many benefits, the undiscerning use of AI tools can have a negative impact on learning. For example, AI misuse can lead to overdependence on AI tools, reduced critical thinking or clinical reasoning, reduced collaborative work, and ultimately reduced patient engagement (6). Additionally, researchers have raised ethical and equity related concerns. Student submissions may meet the institutional criteria for plagiarism or may include factual inaccuracies, and AI tools such as chatbots can share widely their inaccuracies and biases. Furthermore, AI tools such as simulation platforms or virtual reality can be expensive, which can increase the equity gap. This equity gap would have a negative impact on institutes serving underserved populations (7). In a race for institutions to offer "up to date" technology and resources to students in an effort to remain competitive, the guidelines, policies, practices, or even institutional understanding of what is acceptable or unacceptable in terms of academic dishonesty have lagged behind. When a lack of institutional dialogue about how AI shifts student learning, assessment, critical thinking, and faculty roles, exists, the responsibility will land squarely in the individual classroom and on faculty. The central question that emerges from the chaos is, "What tool(s) do individual faculty have to try and bring clarity to the ever-changing situation?" From our perspective, the syllabus becomes the answer. Within it, faculty have the academic freedom to promote AI literacy and understanding, all while trying to deter academic dishonesty, but what does that actually mean? PointThe course syllabus, also known as a course plan, is a multifaceted teaching tool that serves as a guidance document for interactions between instructors, learners, and other stakeholders such as administrators and accreditation bodies. Often, the course syllabus is the first interaction between students and instructors at the beginning of a semester (8). As such, using syllabi to establish norms or frameworks of AI usage for the course would be a logical start.However, given the wide variety in AI tools (e.g., writing tools, image generators, music generators etc.) and evolving nature of the technology, crafting a robust AI framework that serves well for a semester can be a challenge and a heavy lift for faculty. While formulating a syllabus statement for AI usage, we suggest including at least three elements -acceptable use, necessary citations or AI usage disclaimers, and consequences of non-adherence to the policy. See Figure 1. This recommendation aligns with research and scholarly works from other disciplines (9). Here we delve deeper into each of these elements:• Acceptable use: It is important for faculty to determine, and outline, what AI tools and use is acceptable for students as they work through readings, assignments, exams, labs, and learning in the course. Corbin et al (2025) have Previous literature has noted that when instructors do not provide AI-related guidelines, students decide on a case-by-case basis about their AI usage for an assignment, which results in undue cognitive and emotional load. Additionally, instructors need to provide unambiguous, actionable steps related to AI usage (10). These steps can be collectively determined during various professional development seminars, shared repositories, and peer mentoring programs (10). This is also an opportunity for faculty to disclose what AI detection tools they will be using in the assessment of student work (i.e. Turnitin). • Necessary citation: While submitting their assignments, students need to disclose their AI usage through citations. A formal AI citation would include details such as the name of the tool, version number, access date, prompt used, and appropriate links. Professors can permit an informal approach too where AI attributions are made for specific tasks such as image generation, etc. (11). Providing samples of citations in the syllabus can help students understand expectations. Professors can also set an example by disclosing their own AI usage while developing various course related documents. • Nonadherence policy: A syllabus serves as a tool to set unambiguous expectations and information of policies. When students are clearly informed about their rights and responsibilities, it helps in reducing or preventing grievances (8). Hence, including information about possible consequences of non-adherence to the acceptable AI usage and documentation rules, which may include but not limited to resubmission of assignments, a failing grade in the course, or academic probations, would be helpful to students.Additionally, instructors can include institutional grade grievance processes (or links to institutional grade grievance process pages) to support students and ensure due process is followed when grade grievance is reported. By being fully transparent in the syllabus, it can set the groundwork for proper and ethical AI usage in the course and remove some of the AI chaos. As much as we would like to argue that the syllabus is the complete answer to the AI chaos medical education faculty and students are experiencing, we acknowledge it is a first and immediate, step that faculty could take. However, if this is the only step taken, imagine as a student the variety of AI frameworks and rules you might see presented in a single semester. This would make keeping track of all of that overwhelming and may result in more cases of academic dishonesty than before; and unintentional cases at that. It also makes medical faculty the only accountable entity, when the institutional bares some responsibility as well. faculty and students are experiencing, we also acknowledge that, to many, it may be seen as a piecemeal solution and one that puts all the pressure on individual educators. As a student, the variety of AI frameworks and rules you might see presented in a single semester could make keeping track of all of that overwhelming. This could result in more cases of academic dishonesty than before; and unintentional cases at that. An institution-wide policy can help in creating crosscourse consistency regarding norms of AI usage by students. Thus, institution-wide AI policies, guidelines, or frameworks are critical as they serve as a basis for all faculty and students to operate.Broadly speaking, AI course syllabus statements provide guidance about academic integrity and prevent overreliance on AI technology. Institution-wide frameworks are necessary for fostering ethical adoption of AI and bias & misinformation prevention. The elements of AI course syllabus statements and the elements of institutional AI framework are summarized in Table 1. Institution An institute-wide AI framework may be necessary for compliance with regulatory bodies such as accreditation committees or abiding region-specific AI laws such as the European Union Artificial Intelligence Act, and may benefit from a campuswide AI framework (12). Research has indeed highlighted the need for a campus-wide policy for the AI use in medical education with an emphasis on academic integrity and data privacy (13). When medical students use word-processors with embedded AI tools, they may inadvertently share patient data with a technology company. A campus-wide policy or guidance can create awareness about AI usage that is compliant with the Health Insurance Portability and Accountability Act (HIPAA), Family Educational Rights and Privacy Act (FERPA) or similar applicable regulations (14). While FERPA was created in 1974, educational institutions in the US still uphold its principles of fairness and data privacy. Since AI technology is recent, institutions may need human oversight (humanin-the-loop) to implement FERPA principles (15). An institute-wide policy can help in implementing the spirit of FERPA as we usher in an AI in education era.Previous literature has also noted institutional responsibilities for ethical adoption of AI tools in medical education (16). Professors, despite being trained with adequate AI literacy, find it difficult to accurately detect AI-generated assignment submissions. Plagiarism detecting software may not detect AI-generated content accurately or consistently. As such, thoughtfully integrating AI tools and content about AI ethics in the institutional curriculum and courses needs to be the new norm. However, students (and eventually professors) need to be responsible for a critical evaluation of AI-generated output. For example, including assignments to cross-check AI-generated output can train students for responsible use of AI (16). These critical thinking skills can have a long-term impact on health professionals. It will train them to evaluate output from diagnostic or clinical AI tools to prevent misdiagnosis or indiscriminate use of medication. While AI tools can identify clinical diagnosis, their algorithms may have potential biases and diagnostic blindspots. These tools are promising but are prone to errors in diagnostically difficult cases. Given their unreliable diagnostic accuracy, these tools can only be used as an aid to human cognition (17). Institution-wide AI usage framework can play a major role in achieving ethical adoption of AI tools with human oversight.It is easy to say the answer is to develop a campus-wide AI policy and understanding, but implementation could prove to be more challenging. First, institutions are dealing with a variety of AI perspectives, comfort levels, and disciplines/specialties. Stakeholders have to be open-minded and willing to come to the table, which creates an opportunity for increased professional development, retooling, and growth. Secondly, a sense of urgency has to be developed given how often AI tools and technology change, but that can't usurp the grassroots buy-in of medical professionals and faculty.Institutions need to find a balance between hearing all perspectives, understanding concerns, and getting something on paper. This may result in more generalized and vague AI frameworks, as first drafts, to ensure everyone can operate within the framework. As stakeholders engage with the framework, further edits can always be made! Lastly, the institution has a responsibility for ensuring understanding and compliance with all parties, including students. Clear policies and procedures need to be developed alongside the framework that uphold education codes, etc. This is how institutions can protect faculty, be competitive, and still offer a positive learning experience for students. This description of developing campus-wide AI policy builds on adoption trajectory of AI (enthusiasm, uncertainty, pragmatic implementation, and thoughtful integration) as described by Izquierdo-Condoy et al (2026) (18). As we consider the point (AI usage statements in the course syllabus) and the counterpoint (institution-wide AI usage framework), we recognize that there is a degree of overlap in these frameworks. Both frameworks need to foster ethical usage and transparency from students and professors. Additionally, both frameworks need to uphold the highest standards of academic integrity to promote learning.Yet these frameworks are not a substitute for each other. While institutional frameworks are crucial for campus-wide ethical adoption of AI, it may not be specific enough to meet student learning objectives (SLOs). For instance, an institute-wide AI policy may have a broad statement such as 'the use of AI writing tools may be permitted by instructors as long as no patient data is shared'. However, physiology instructors may choose to not permit AI writing tools in the course as one of the SLOs iwas 'to demonstrate proficiency in medical communications through scientific writing'. The course AI policy would fall within the parameters of the institute-wide framework. On the other hand, if a physiology faculty member allows AI writing tools, but their syllabus policy statement doesn't address the restriction of sharing patient data it would not be aligned to the institute-wide framework and open any student sharing that information on the AI tool, the faculty, and the institution to a potential lawsuit. AI usage syllabus statements and an institution-wide AI usage framework need to complement each other. When instructors observe non-adherence to AI syllabus statements, a punitive action may be taken. These consequences need to align with campus-wide academic dishonesty policies, e.g., students receiving a failing grade or academic probation.Given the complexity of the AI tools, we recognize that instructors may be hesitant to write their AI usage course syllabus statements. Instructors need to be supported via various professional development seminars, syllabus generator tools, and support from instruction designers. These frameworks need to be reviewed periodically as technology continues to evolve.While syllabus statements and campus-wide AI frameworks can offer immediate and short-term solutions, these documents will need to be reviewed periodically as the technology continues to evolve. As such, further research (or institutional evaluations) related to the effectiveness of three element syllabus statements and student perspectives about these documents need to be conducted. Future directions can also include revamping assessment methods to make them more inclusive of AI tools and resistant to AI usage for cheating purposes. In summary, we believe that physiology professors can thoughtfully and ethically integrate AI tools in their curriculum by adding AI usage course syllabus statements. While the institution-wide frameworks can serve as a broader AI usage framework, these course syllabus statements can guide effective and efficient use of AI tools.
Keywords
AI framework, AI policy, Course syllabus, Institutional framework, physiology education
Received
19 February 2026
Accepted
15 May 2026
Copyright
© 2026 Karve and Hurless. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence:
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.