How Physiology Educators Are Adapting to Artificial Intelligence

As artificial intelligence enters the classroom, physiology educators explore how to use new tools responsibly while rethinking assessment and instruction.

By Suchi Rudra

As the world grapples with the unstoppable integration of artificial intelligence (AI) into everyday life, physiology educators are examining how it can be incorporated into their teaching, while staying mindful of concerns about ethics and reliability. AI has been around for decades—it’s generative AI that is the relative newcomer, causing confusion and some distrust, as any new technology tends to do. Generative AI is a type of AI that generates content, such as text, images, audio and video. The content is not original but created when a large language model responds to prompts and questions by “predicting” the content based on the model it’s trained on.

Recent surveys show that between 80% to 90% of college students in the U.S. are regularly using generative AI tools. But what about their professors?

Lina Shehadeh, PhD, a cardiovascular research professor at the University of Miami Miller School of Medicine, began to offer a twice-a-month AI workshop this academic year for the school’s students and faculty. Her main goal was to educate the PhD students in the school, so she was surprised when many faculty members—including those in leadership positions—began to attend. That showed her there was serious interest in getting more clarity around the use of AI.

In the workshop, Shehadeh likes to start by saying, “We understand the discomfort and anxiety. We know this is a big issue.” That tends to calm everyone down, she says. “If we encourage them and approach them from this angle, and say, ‘Listen, it is doable; you don’t need prior experience’ and that we can define what’s right and what’s wrong when using AI, I think you win them. Everybody wants to learn; everybody wants to be a better teacher,” she says.

Any hesitation, especially from students, tends to be about ownership: If AI creates content for you, is it still your work? Shehadeh says yes. Part of what she emphasizes in her workshop is the necessity and importance of disclosures when using AI, which can help put the user at ease. “You can say it’s yours, but acknowledge that you used AI to generate the [content]. It’s just a method you used. If you used AI to edit your writing, say it; there’s nothing wrong with it,” she says.

Becoming AI Literate

Clearing up these kinds of questions is part of AI literacy, which has become a critical part of a college education—for both educators and students. At Gettysburg College, Josef Brandauer, PhD, dean of academic advising and student success, insisted on having his students use AI for certain assignments. “I make them use it because I want to know that they know,” he says.

Before becoming a dean last year, Brandauer was a Gettysburg College professor of health sciences for 17 years. In his physiology courses, he had his students use generative AI to answer a question, which would give them “a very polished, probably excessively wordy and confident answer.” The real task was to evaluate the accuracy of the answer: Was it true, false, complete? Was it missing information or context?

“That’s a really, really important skill for any student to obtain, whether AI exists or not. We need to not be confidently wrong, and we can use AI as a tool by which we practice those skills that are important to us,” Brandauer says.

Although the onslaught of AI platforms feels relatively new, Brandauer compares this moment in time to the introduction of the graphing calculator and statistical software programs. “We are teaching statistics differently now than we did 25 years ago. We now have another tool that can be highly effective. One of the responsibilities that we have is to help our students understand how those tools work, ethically, ecologically soundly, and what their flaws and shortcomings are,” he says.

Adjusting Assessments

While students have embraced AI, educators are also beginning to use it in their course planning. Tools include AI writing assistants, productivity tools and coding agents that can create apps. Mastering these tools could save time and energy—especially when it comes to creating and grading assessments—and transform the teaching experience.

The University of Miami medical school, for example, has a committee dedicated to reviewing all assessment questions for every course and using AI to “fix and perfect” them in collaboration with the professors, Shehadeh says.

Some educators are reacting to students’ use of AI by changing their assessment practices to help make sure students are still doing the work. Brandauer says he and his department colleagues changed their assessment structures “for the better, and AI was probably a contributing factor to that.” In his cardio-respiratory class, Brandauer made the decision to switch to an oral final exam, where each student gave a 20-minute teaching demonstration with a visual aid of how oxygen makes it from the atmosphere to the respirating mitochondria, plus a 10-minute Q&A. “There is absolutely no way that they can use AI on the spot, and AI doesn’t really help them all that much in the presentation,” he says.

George Kararigas, PhD, professor of physiology at the University of Iceland’s Faculty of Medicine, is part of an international consortium testing AI tools in education. He is looking forward to soon testing a third-party AI-driven assessment tool that was developed to grade even essay-based exams, in addition to multiple-choice tests. However, while AI tools might save time, he says, there’s also the pain point of constantly learning to use new tools, a process that can be time-consuming and challenging.

Saving Time

Indeed, one problem all physiology educators share: never enough time. It makes sense then that the most attractive AI feature is how much time a tool can save. At University of Miami’s medical school, Laura Bianchi, PhD, professor of physiology and biophysics, explains one medical school challenge: Basic sciences are very condensed at the beginning of the curriculum and students are immediately thrown into clinic. “Because the basic science of physiology training is so shortened, there have been concerns that we need to bring back some of these concepts later in that curriculum, in Phase 2 and 3, when they are doing clinics. But it’s a huge task,” she says.

Enter AI. Bianchi and Shehadeh collaborated on a project to map existing resources of physiology medical licensing exam-like questions onto all the clinical clerkships in Phases 2 and 3. By using AI to scan the questions and curriculum and then suggest the mapping points, the previously insurmountable task was made possible. Without any coding expertise, Shehadeh used Replit, an AI-powered software program, to develop an interactive educational module in several days instead of several months. The module offers a refresher of the basic concepts covered in the United States Medical Licensing Examination Step 1 exam. Shehadeh says there’s an urgent need across the country to create these mini lessons on foundational science, which can help students integrate basic science knowledge into their clinical training.

The module, with content supported by a podcast and simulations, has now been available and active for the past academic year, and so far, student feedback is positive. “They don’t just want the same stuff, the same material as when they were first taught the concept,” Shehadeh says.

At this point, Shehadeh has created one module for one basic science concept. She says the other 44 still need to be developed. “It would be great if this could be a shared project among other universities as well. This is an enormous task. It’s across so many schools in the nation; it’s not just us. I call for collaborators. Why don’t we have some from other universities and make this a public physiology miniseries? We want to make it available for everyone—for underprivileged students too.”

Perhaps educators will turn more readily toward the use of generative AI when it can be used to solve a current challenge, like the creation of the refresher modules. But in any case, it’s only a matter of time before AI is fully integrated into every physiology classroom, experts say. So it helps all educators to be ready, to be proactive.

“The worst you can do is ignore AI,” Brandauer says, “because it’s out there and people are using it. When we used to start doing workshops on my campus and around the country, the people that were most worried about AI generally were the folks who hadn’t engaged with it at all. Just taking that first step is important, and then from there, keeping up with the technology and really knowing what its capabilities are is a really important piece.”

This article was originally published in the January 2026 issue of The Physiologist Magazine. Copyright © 2026 by the American Physiological Society. Send questions or comments to tphysmag@physiology.org.