Pharmacists are embracing AI to handle routine medication questions and administrative tasks, but a growing consensus among healthcare educators emphasizes that clinical judgment must stay in human hands. At the University of Rhode Island College of Pharmacy's 41st annual Seminar By the Sea conference, held March 19-20, educators embedded AI training throughout the program to help pharmacists understand both the promise and the pitfalls of integrating these tools into everyday practice. Why Are Pharmacists Getting Hands-On AI Training Now? Rather than treating AI as a futuristic concept, the URI College of Pharmacy designed interactive sessions that reflect real-world medication management workflows. Participants explored practical applications including drug information retrieval, patient communication with language translation, and administrative efficiency improvements. The conference demonstrated how AI can reduce documentation burden while supporting clinical decision-making, but with a critical caveat: the technology must always operate under human oversight. Mary-Jane Kanaczet, director of continuing professional development at URI, explained the philosophy behind the approach: "These conversations are woven throughout the conference as part of our focus on translating emerging tools into safe, effective, and patient-centered medication use." The goal is not to replace pharmacists but to free them from routine tasks so they can focus on complex clinical judgment. What Happens When AI Gets the Dosage Wrong? During one interactive session, audience members tested an AI platform by asking health-related questions, including whether combining the antibiotic Linezolid with a fentanyl infusion posed a safety concern. While the AI provided clinically sound advice to monitor patients carefully, it included an important disclaimer: "I am not a physician or pharmacist". This distinction matters enormously. Sean Jeffrey, professor of pharmacy practice at the University of Connecticut and director of pharmacy at Hartford Healthcare Group, presented research on the progression from today's narrow AI systems toward Artificial General Intelligence (AGI), where machines begin learning independently, and eventually Artificial Super Intelligence (ASI), where machines exceed human capability. Some scientists believe ASI could arrive within four years, which makes establishing safeguards urgent. "We're moving closer to clinical services being performed without a human, but what happens when AI gets a dosage wrong? We need to keep humans in the loop," Jeffrey stated. He emphasized that pharmacists must retain ultimate decision-making authority, even as AI tools become more sophisticated. How to Implement AI Safely in Pharmacy Practice - Prompt Refinement: Train pharmacists to ask AI tools better questions and understand how to structure queries for clinically useful, reliable outputs rather than accepting initial responses uncritically. - Verification Protocols: Establish workflows where AI recommendations are reviewed by licensed pharmacists before implementation, ensuring human judgment validates all clinical decisions. - Transparency Requirements: Demand that AI systems disclose their limitations and confidence levels, similar to how the tested platform acknowledged it was not a licensed healthcare provider. - Continuous Monitoring: Use AI for patient surveillance and early warning systems, but require pharmacists to interpret results and determine clinical significance. - Documentation Support: Deploy AI to reduce administrative burden through automated transcription and record-keeping, freeing pharmacists to spend more time on direct patient care. Kerry LaPlante, dean of the URI College of Pharmacy, reinforced this philosophy: "AI has tremendous potential to enhance how we deliver medication-related care, but it must be approached thoughtfully. Our responsibility as a college is to prepare pharmacists who can critically evaluate these tools, integrate them safely into practice, and always keep patient care at the center". What Real-World AI Applications Are Already in Use? Healthcare organizations are already deploying AI in clinical settings. Ambient listening technology automatically transcribes patient-physician interactions during appointments, reducing documentation time. Companies like MedMe Health offer pharmaceutical chatbots that answer patient calls, gather information, book appointments, and respond to routine questions. Wearable technology monitors vital signs continuously, and diagnostic platforms help identify symptoms. Beyond medication management, precision oncology is advancing through AI-driven insights. Caris Life Sciences announced new AI signatures that predict the risk of brain metastases in breast and lung cancer patients. These proprietary algorithms, trained on data from 12,994 non-small cell lung cancer cases and 3,371 breast cancer cases, generate personalized risk scores using whole exome sequencing and whole transcriptome sequencing data. "Insights from these proprietary Caris AI signatures give us a forward-looking view of which patients may be at elevated risk for brain metastases, allowing us to help guide clinicians to shift from passive surveillance to more proactive monitoring," said David Spetzler, president of Caris Life Sciences. David Spetzler, President, Caris Life Sciences The Caris MI Cancer Seek assay received FDA approval in November 2024 as the first and only simultaneous whole exome and whole transcriptome sequencing-based assay with FDA-approved companion diagnostic indications for molecular profiling of solid tumors. Results are visualized as Kaplan-Meier curves, providing clinicians with an intuitive view of likelihood and rate of brain metastasis development based on each patient's molecular profile. Why Does Trust in AI Matter More Than the Technology Itself? Jeffrey emphasized a fundamental truth that extends beyond pharmacy: "The tech doesn't matter if people don't trust it." Organizations like the Coalition for Health AI are working to establish standards for responsible adoption, recognizing that clinician confidence is essential for widespread implementation. Jeffrey The challenge is not whether AI can perform certain tasks better than humans. In many cases, it can. The challenge is ensuring that healthcare systems maintain human oversight, accountability, and the ability to catch errors before they harm patients. As AI systems become more capable, the stakes of getting this balance wrong increase exponentially. The pharmacy profession's approach to AI education offers a model for other healthcare disciplines. Rather than resisting the technology or adopting it uncritically, educators are teaching clinicians to view AI as a powerful tool that enhances human expertise rather than replaces it. This philosophy keeps patients at the center of care while allowing healthcare professionals to work more efficiently and make better-informed decisions.
