When the editors of the Journal of General Internal Medicine opened the submission portal for their eighth Medical Education Theme Issue, they received 113 manuscripts and accepted just over 25 for publication. The resulting collection, assembled with the help of 12 associate editors specializing in medical education, arrives at a moment of unusual turbulence for the field. Artificial intelligence has swept into classrooms, clinics, and clerkships faster than the evidence base needed to guide its use, and the 2027 Theme Issue, described in an editorial published on 23 September 2026, positions itself as both a snapshot of that disruption and a bridge to two decades of prior scholarship. Since the first Theme Issue appeared in 2004, the series has been published every two to five years and has highlighted more than 220 manuscripts, making it one of the longest-running barometers of how physicians are trained in general internal medicine.
The editorial team, led by Stephen Fuest of the University of Colorado School of Medicine, Nora Osman of Brigham and Women’s Hospital and Harvard Medical School, and Wei Wei Lee of the University of Chicago Pritzker School of Medicine, did something unusual in preparing their retrospective: they went back to the original May 2004 issue and read it closely. What they found was a field wrestling with a technology that now seems almost quaint. The 2004 issue carried multiple manuscripts on the impact of the World Wide Web on medical education, including new online curricula and web-based clinical skills evaluation systems. Another cluster of papers examined the consequences of newly imposed resident duty hour limits, including the 80-hour workweek and 24-hour consecutive shift caps, on patient experience. Twenty-three years later, the editors noted, duty hours matter less to patient experience than they once did, but the technology theme has returned with startling force.
One manuscript from 2004 received particular attention in the new editorial. Titled “Developing a Peer Review Process for Web-based Curricula,” it documented the Society of General Internal Medicine’s method for assessing web-based education tools using a structured peer evaluation instrument. The authors reviewed the literature on distance learning, internet curricula, the differences between print and web formats, and learner evaluations of online tools, then applied their instrument to 15 abstracts and websites submitted for the Innovations in Medical Education presentations at the 2003 SGIM meeting. The editors highlighted the study for three reasons. First, it is a striking reminder of how thoroughly the internet has since permeated clinical care and education, to the point that a day in general internal medicine rarely passes without it. Second, the manuscript is both similar to and different from today’s AI moment: educators were then collaborating to establish best practices for a new technology, but the internet of 2004 had far different capabilities than the generative systems of 2027. Third, JGIM has already become a home for AI-related education publications, and SGIM has proactively explored best practices and emerging challenges through its AI Task Force and a formal Position Statement.
That position statement, published in JGIM in 2025, offered recommendations for clinicians, technologists, and healthcare organizations on the use of generative artificial intelligence in medicine. The editorial’s authors build on that foundation with a clear stance: they believe AI will augment, not replace, educators in general internal medicine. Frameworks and strategies already exist that can help educators incorporate the technology while identifying standards for competency, including recent work in the New England Journal of Medicine on educational strategies for clinical supervision of AI use. Still, the editors indulge in a wry thought experiment: perhaps in 25 years, future educators will write a reflection piece in JGIM on the ubiquitous nature of AI manuscripts in 2027, much as today’s editors marveled at the internet papers of 2004.
The 2027 Theme Issue was built around a deliberate solicitation strategy. The editors invited submissions on AI-related issues in medical education; care delivery innovation involving learners; teaching the value of primary care and supporting future generations of generalists; culture, learning environment, well-being, and psychological safety; and professionalism and professional identity formation. Once the accepted manuscripts were assembled, a somewhat different set of themes emerged organically: transitions across the training continuum; workforce development in general internal medicine; educational innovations; finances in graduate medical education; and continuity clinic. The mismatch between what was solicited and what surfaced, the editors suggest, demonstrates the breadth and unexpected direction of scholarship in the field. Every manuscript was screened against four questions: Is it new? Is it true? Will it change what we do? Is the topic right for the JGIM audience? For the accepted papers, the editors believe the answer to all four is yes.
Several individual contributions stand out. A 2024 survey of the Clerkship Directors of Internal Medicine provides an updated picture of the clerkship landscape, covering structure, duration, curricular models, and the balance of inpatient and outpatient experiences. Several manuscripts address graduate medical education directly, including a descriptive study of intern orientation, research on in-basket management and inter-visit care, and an analysis of resident application characteristics associated with a successful subspecialty fellowship match. The issue also examines physician attitudes toward unionization, a topic with obvious implications for the workforce themes running through the collection. Perhaps most emblematic of how learning itself is changing, the issue describes the development and success of a podcast driven and produced by medical students, illustrating the evolving ways learners engage with and acquire knowledge outside traditional lecture halls.
The editors are candid about the economics of the field. Medical education scholarship often lacks straightforward mechanisms for funding, protected time, or institutional resources, and the effort of all who submitted manuscripts is acknowledged as significant. To improve the odds for future submissions, the editorial spells out the most common reasons manuscripts were rejected: limited relevance to a general internal medicine audience, inadequate positioning of the work within prior literature, small sample sizes, and outcomes limited to Kirkpatrick Model Level 1, the lowest tier of the widely used evaluation framework, which measures only learner satisfaction and reaction rather than changes in learning, behavior, or patient outcomes. The advice is practical and pointed: situate new work against what is already known, design studies with adequate statistical power, and aim for outcome measures that demonstrate genuine educational impact.
In a move that mirrors the very subject it examines, the editorial team used AI to augment their own review of prior Theme Issues. They personally verified the output and reported satisfaction with the efficiency and thoroughness of the process. The technology, they write, allowed them to take a deeper dive into prior issues by removing the operational, time-consuming task of categorizing papers and identifying evergreen themes, freeing them to focus on the deeper cognitive work of meaning-making and writing. The disclosure section of the editorial notes that the authors were solely responsible for conceptualization and drafting, with OpenAI’s ChatGPT large language model employed for final editorial assistance to enhance grammar and readability, generate title ideas, and help identify prior Theme Issues after the authors manually searched JGIM volumes. The editors describe themselves as cautiously optimistic and curious about how AI will affect medical education going forward.
The editorial’s most substantive analysis concerns the widening gap between AI adoption and AI evidence. The technology has already altered medical education in ways ranging from how easily learners access information to how tasks that once required significant human effort are delegated. AI can help learners and physicians gather and synthesize information, generate and refine differential diagnoses, and may support clinical reasoning. For educators, it can assist with assessment, particularly the difficult work of synthesizing multiple observations into useful formative and summative narrative evaluations. Yet, the editors warn, understanding has not kept pace with adoption; enthusiasm may be years ahead of both preparedness and the evidence that would normally guide adoption of a new intervention in clinical medicine.
The stakes of that gap extend beyond task accuracy. Medical education, the editors argue, is fundamentally interpersonal, interactive, and iterative: learning happens through dialogue with teachers and peers, direct patient care, observation, feedback, and the cumulative development of critical reasoning. If AI augments those interactions, it may prove enormously valuable. If it replaces them, the risks include isolation, overreliance, propagation of errors, and erosion of the very processes through which clinical judgment develops. Unresolved ethical questions also shadow both medical education and patient care. For now, the editors conclude, much of this landscape remains unexplored, and they hope future Medical Education Theme Issues will carry that exploration forward, approached with humility and openness and an honest acknowledgment that the consequences of widespread AI use are not yet understood. The closing thanks go to the journal’s co-editors-in-chief, editorial staff, and medical education associate editors, whose shared effort, the authors believe, will leave readers better informed and inspired to keep growing the collective knowledge base of medical education.
Subject of Research: The 2027 Journal of General Internal Medicine Medical Education Theme Issue and the role of artificial intelligence in medical education
Article Title: From Evidence to Education: New Insights in the 2027 JGIM Medical Education Theme Issue
Article References: Fuest, S., Osman, N., & Lee, W. W. (2026). From Evidence to Education: New Insights in the 2027 JGIM Medical Education Theme Issue. Journal of General Internal Medicine. https://doi.org/10.1007/s11606-026-10817-5
Image Credits: AI Generated
DOI: 10.1007/s11606-026-10817-5
Keywords: medical education, artificial intelligence, general internal medicine, JGIM, residency training, graduate medical education, clinical reasoning, curriculum, assessment, generative AI, workforce development, continuity clinic
Cite Scienmag News
APA MLA Chicago
Courtney Benton. (September 24, 2026). AI Takes Center Stage as JGIM Marks Two Decades of Medical Education Scholarship. Scienmag. https://scienmag.com/ai-takes-center-stage-as-jgim-marks-two-decades-of-medical-education-scholarship/
Courtney Benton. “AI Takes Center Stage as JGIM Marks Two Decades of Medical Education Scholarship.” Scienmag, 24 September 2026, https://scienmag.com/ai-takes-center-stage-as-jgim-marks-two-decades-of-medical-education-scholarship/. Accessed 24 September 2026.
Courtney Benton. “AI Takes Center Stage as JGIM Marks Two Decades of Medical Education Scholarship.” Scienmag. September 24, 2026. https://scienmag.com/ai-takes-center-stage-as-jgim-marks-two-decades-of-medical-education-scholarship/
Copy citation Download RIS
Tags: AI integration in medical classroomsArtificial Intelligenceassessmentchallenges of AI adoption in clinical trainingclinical reasoningcontinuity cliniccurriculumevolution of internal medicine trainingfuture of medical training with AIgeneral internal medicinegenerative AIgraduate medical educationhistory of medical education themesimpact of artificial intelligence in healthcareinterdisciplinary collaboration in medical educationJGIMJournal of General Internal Medicinelong-term trends in medical scholarshipMedical Educationmedical education research milestonesMedical education scholarshipresidency trainingscholarly publication trends in medicineWorkforce development
