Artificial Intelligence in Pharmacy Residency Training: Striking the Right Balance

Artificial intelligence is advancing at a rapid pace and is predicted to exponentially grow as time progresses. In healthcare, and particularly in pharmacy, AI has moved well beyond the hype stage and into practical, everyday tools that are actively reshaping clinical practice and education. As pharmacists, we are already interacting with these technologies daily, and the pace of integration continues to accelerate. The real question for us now is no longer whether AI belongs in pharmacy, but how do we intentionally prepare the next generation of pharmacists to use it wisely throughout their residency training, knowing this is exactly the environment they will enter after graduation.

Today’s pharmacists are leveraging AI across clinical, operational, and administrative domains. Ambient listening platforms capture patient encounters in real time and generate clinical notes to reduce documentation burden.¹ Literature trained clinical evidence platforms and tools built on large language models (LLMs), such as OpenEvidence, are used daily by almost half of U.S. physicians and supported tens of millions of clinical consultations in December 2025.² Pharmacists use it to deliver rapid, literature-backed summaries of the latest research, replacing hours of manual literature searching for medication questions. Behind the scenes, AI streamlines patient access like prior authorizations, screens for patient financial assistance eligibility, and integrates directly into electronic health record workflows to minimize care delays.^3,4 These tools are actively used in health systems right now, delivering measurable gains for pharmacists.

For pharmacy residents, this shift creates both tremendous opportunity and educational risks. When used thoughtfully, AI acts as a powerful learning accelerator, handling mundane and repetitive tasks so residents can focus on higher-level clinical reasoning, complex patient cases, research activities, and direct patient care. It improves efficiency in educational and scholarly work while still allowing time for deep skill-building that defines residency. Current uses include decreasing administrative tasks (i.e. summarizing meeting minutes, generating emails, and finding clinical articles for knowledge management), improving project management (i.e. creating accurate timelines and breaking down tasks into manageable steps), and increasing preparation for residency and post-residency employment (i.e. preparing for job interview questions and reviewing/improving application materials). LLMs can also help residents build structured study plans for rotation topics, generate mock topic discussion questions, compare treatment guidelines across disease states, rehearse patient counseling points, and outline presentations or teaching sessions before refining them with preceptor feedback. In research and teaching, AI can help organize background sections, identify gaps in early drafts, translate dense content into patient-friendly language, and create first-pass summaries that residents then verify against primary sources. By using synthetic patient prompts, LLMs can simulate patient encounters in a safe practice space, giving residents an opportunity to sharpen communication skills and reinforce key clinical pearls during comprehensive medication reviews. Used this way, AI supports self-directed learning and communication while keeping the resident accountable for judgment, verification, and final clinical decisions.

However, over-reliance carries genuine dangers. If AI becomes a crutch for core activities like literature appraisal, care-plan development, or clinical decision-making, residents risk missing the foundational competencies that residency is designed to instill. Previous research has shown that overreliance on AI decreases students’ ability to think critically, reason analytically, and make independent decisions, primarily because they may adopt AI output uncritically and favor quick solutions over deeper engagement.⁵ Interpreting primary literature, developing independent critical thinking, and feeling comfortable navigating clinically ambiguous situations are key for residents to develop as they prepare themselves for advanced clinical practice. Even more concerning, if residents never traditionally learn how to do these tasks, they lose their ability to effectively judge the quality of AI outputs, recognize when the tool is inaccurate, or spot its common pitfalls and limitations. The American Society of Health-System Pharmacists released its Statement on Artificial Intelligence in Pharmacy in August 2025, calling on pharmacists to lead in the design, implementation, governance, and safe use of these tools.⁶ Due to the rapid development and advancement of AI technologies, pharmacists need to be on the forefront as both clinical and educational activities can be significantly impacted.

There are also practical hurdles residency programs must address: safeguarding patient privacy, integrating AI smoothly into existing workflows, training faculty with varying levels of familiarity, and building informatics and technology skills across the team. Looking ahead, we are preparing residents who belong to generations already far more comfortable with AI than previous ones, just as Microsoft Excel and PowerPoint became expected competencies after their introduction to society. AI proficiency will soon become a baseline job requirement, yet significant gaps remain in the literature leading to limited formal recommendations around what’s considered useful versus inappropriate use of the technology.

AI could change how we assess residency applications and scholarly projects, so programs should decide in advance where AI-supported work is acceptable, where disclosure is required, and where independent performance must still be demonstrated. For residency applications, letters of intent should likely carry less weight as stand-alone writing samples and instead be paired with structured interview questions that test whether applicants can clearly explain, personalize, and defend what they submitted. For virtual interviews, programs should set explicit expectations around AI use and rely more heavily on behavioral follow-up questions, real-time discussion, and requests for candidates to explain their reasoning rather than simply provide polished answers. Interview case presentations should emphasize patient-specific ambiguity, evolving clinical details, and on-the-spot defense of recommendations rather than topics that can be addressed with a quick AI search. For scholarly work, AI can reasonably assist with administrative drafting tasks such as formatting or refining an IRB protocol, but residents should still be expected to understand and defend the study question, design, methods, and ethical considerations before submission. Similarly, AI may help accelerate statistical analysis, but residents should demonstrate a working understanding of why a method was chosen, what assumptions apply, how results should be interpreted, and where errors or limitations may exist. Establishing these expectations early helps programs, preceptors, and residents stay aligned on responsible use.

AI is clearly the future of our profession. Rather than passively adopting these tools, pharmacists must actively shape, evaluate, and lead their responsible integration to enhance patient safety and professional development. That type of leadership starts with how we train our residents. Outright prohibition of AI during residency is neither realistic nor helpful. It would leave trainees unprepared for modern practice. At the same time, completely unrestricted use is equally unwise. Residency programs must develop clear, thoughtful frameworks for responsible AI integration. This includes transparent documentation of when and how AI is used, targeted education on its limitations such as hallucinations, bias, and lack of context, and deliberate exercises that still require residents to demonstrate unassisted mastery of essential skills. These frameworks are not only essential for resident learning, but also for determining expectations for program preceptors and clinical staff.

When done well, this balanced approach empowers residents to harness powerful AI tools confidently today while still developing the irreplaceable human expertise needed for excellent pharmacy practice tomorrow.