Prescriptions for Progress: How Pharmacy Education is Using AI

As a first-year pharmacy student, Malena McKinney knew basic coding but had only passing interest in artificial intelligence (AI). Then in two years at Belmont College of Pharmacy & Health Sciences in Nashville, Tennessee, McKinney used AI tools to develop study guides and analyze research data.

That’s taking time. There are not yet any established AI-related competencies for student pharmacists, and faculty members themselves are new to the technologies. Most schools and colleges have just begun laying the groundwork for implementing AI in their curriculums.

“It’s evolving so rapidly that it does create challenges,” said Alan Zillich, a professor at Purdue University College of Pharmacy, who is on a one-year sabbatical studying AI integration into practice and pharmacy education. “What do we need to teach? What do they need to know?”

Institutions recognize that grappling with the technology’s implications is critical to their students’ future practices, said Kaitlin Alexander, a clinical associate professor at University of Florida College of Pharmacy, who last year won the campus-wide inaugural AI Teaching Integration Award.

“We want to ensure they (students) have at least general knowledge and understanding and abilities to apply AI tools when they graduate,” she said.

Teaching students to be competent developers and wise users
Belmont’s pharmacy college had a head start on many institutions because it established an informatics track in 2009, said associate professor Anthony Blash. Experienced as a professional pharmacist informaticist at the corporate level, Blash said that the marketplace needs pharmacists who are comfortable analyzing complex data, the basis for AI. 

At Belmont, students are required to take multiple courses incorporating informatics and analytics in the PharmD curriculum; a pharmacy skills series covers not only compounding and calculations but incorporates data analytics into the student project work across multiple semesters.

“The competent clinician has to be able to manage big data,” said Blash. The college has also developed a technical writing course to help students understand when to use available AI models. The course helps students reflect on the potential uses – and misuses – of AI programs used for specific tasks, such as developing drug information and patient communications.

Students’ projects illustrate the rapid evolution of AI. In 2019, Belmont students developed an Alexa app that provided patients with information on the top 200 prescribed drugs. More recently, their work has included creation of custom ChatGPT-powered models that tutor students on pharmacy-related math problems and track drug shortages. McKinney, the student, created an AI model to develop a lab value interpretation tool for practitioners and an oncology drug study guide for students.

“AI can be a crutch,” Blash said. “If you don’t understand the intent of the AI model, you can think the computer is all-knowing. We poke at that and work to develop robust quality assurance mechanisms for AI.

“We’re teaching our graduate PharmD students to be competent as AI developers and to be able to function in healthcare as competent and ethical users of artificial intelligence.”

Using AI as a step in the workflow
Much of the work with AI in higher education remains the initiatives of individual professors, like Kyle Hultgren. A clinical assistant professor of pharmacy at Purdue, Hultgren said that so many students were discussing ChatGPT two years ago, that he decided to have the class re-do one of their old assignments using AI tools.

“I just wanted to see what it would look like,” he said. “That was my first tip of the toe in the water.”

Hultgren is participating in a university-wide initiative to create a master’s degree in AI, contributing ideas for building courses for healthcare professionals. In his classroom, he discusses AI as part of students’ workflow, such as using AI tools to come up with retail space costs for developing business models.

“Something the students are learning is to integrate these, rather than use them as a substitute for the initial work,” said Hultgren. His colleague, Zillich, noted that educators are working on basic principles, including not only how to use the large-language models but also the benefits and ethical pitfalls.

“So knowing if it’s OK to use it, and what questions you need to ask,” Zillich said. “At the end of the day, whatever you signed off on, it’s assumed you vetted it.”

Augmenting, not outsourcing
Alexander, who participates in a university-wide faculty learning group on AI, was honored by the UF Center for Teaching Excellence for her integration of AI into her advanced pharmacy practice experience critical care elective.

In that course, students typically had been asked to complete drug information questions. She revised the assignment, asking them to develop a prompt for an AI tool to generate an output, which they evaluated before doing their own drug information search. “So they were using AI as a starting point,” said Alexander, “then building on it with their own information to come up with the final submission.”

Alexander is leading a college task force to create a gap analysis determining what content is already being taught and what is missing. She’s also studying new medical education recommendations to see what could apply to pharmacy schools.

UF has also had some faculty education on AI. “We have a large preceptor base, and they all have varying levels of AI knowledge,” she said. “This fall, we did an introduction to AI for preceptor development. Our preceptors teach the final year so we want to know what students are getting on their experiential rotations.”

Another idea that excites her is building more AI-specific applications to help students study. “We’ve seen a lot of interest from students in learning more about that, whether it’s giving them specific prompts or building tutoring chat bots. We don’t want students to outsource their problem solving to these tools but (instead) ask how we can use the technology to augment learning. We’re trying to be intentional so we can get them knowledgeable but still hold them accountable.”

Teaching AI as a step in the thought process
At the University of Pittsburgh School of Pharmacy, Ravi Patel serves as professor and lead innovation advisor. In that role, he says, he looks for useful solutions in healthcare and helps connect students with health systems and technology start-up companies.

For example, the college developed a course that featured input from private company leaders, who discussed how pharmacy can help shape future workflow operations. “Clinician involvement in the generation of solutions is a key potential that I hope pharmacy has a chance to realize,” he said.

Patel said helping students develop their critical thinking skills around AI is essential, even if pharmacy licensing exams don’t measure those sorts of skills. “How do you standardize assessing thought process and problem solving in something like a multiple-choice exam?” he said. “We can look toward technology as thinking of problem-solving approaches rather than right answers.”

Twenty years ago, Patel noted, being able to use a Palm Pilot to reference a drug resource would have been a valuable skill. By 2015, creating a PivotTable in Microsoft Excel was critical to use many medical technology platforms.

“Today, in 2025, being able to create a drug information site from scratch using generative AI is as easy as those other things,” he said. “But the thought process that goes into solving problems has been consistent.”