Generative artificial intelligence is rapidly entering medical education. Large language models, virtual patients, adaptive learning systems, AI-assisted assessment, and medical education data analytics are beginning to reshape how clinical reasoning, teaching feedback, and learning support are designed.
This shift is also changing what medical teachers are expected to do. In addition to traditional roles such as knowledge instruction, clinical supervision, and learner assessment, teachers are increasingly expected to judge whether AI-generated content is reliable, design human–AI collaborative teaching activities, manage educational and clinical data responsibly, and guide learners in understanding algorithmic bias, model hallucination, privacy protection, and ethical boundaries.
Against this background, the China Consortium of Elite Teaching Hospitals has published the Consensus on the Digital Intelligence Competency Framework for Medical Teachers in the Chinese Journal of Medical Education Research . Led by the consortium and implemented through the “Future Medical Education Leadership Initiative” at The First Affiliated Hospital, Zhejiang University School of Medicine, the consensus was developed with input from medical education experts, digital medicine specialists, artificial intelligence experts, teaching administrators, and frontline clinical faculty.
The article responds to a practical question facing medical education: in the era of generative AI, what competencies should medical teachers have?
The consensus argues that medical teachers need more than the ability to use AI tools. They need the capacity to understand the limits of intelligent technologies, supervise their use critically, manage risks, protect patient privacy, and integrate AI into teaching in ways that remain educationally meaningful and ethically responsible.
The framework uses the term “digital intelligence competency” to emphasize this broader skill set. It refers not only to digital literacy or technical operation, but also to the ability to combine digital tools, intelligent systems, ethical judgment, and teaching innovation. The goal is not to turn every medical teacher into an AI engineer. Rather, it is to help teachers use AI appropriately in medical education, recognize its risks, and maintain their central role in clinical reasoning training, professional formation, and humanistic education.
The consensus identifies five core competencies for medical teachers in the GenAI era: digital intelligence knowledge foundation, digital intelligence application skills, digital intelligence ethics and security, digital intelligence teaching integration, and digital intelligence research translation. These five areas are further divided into 21 secondary indicators, forming a tiered structure that moves from foundational capabilities to advanced application and higher-level innovation.
The first competency, digital intelligence knowledge foundation, requires teachers to understand the basic concepts of generative AI, large language models, prompts, tokens, pre-training, multimodal integration, and the use of AI in medical education. It also emphasizes awareness of the limitations of these technologies, including data bias, outdated knowledge, misunderstanding of professional terminology, and hallucinated outputs.
The second competency, digital intelligence application skills, focuses on practical use. Medical teachers should be able to apply prompt strategies, prepare or process medical data appropriately, and identify and correct errors in AI-generated content. The consensus highlights that teachers should not directly use AI outputs in teaching without verification. Instead, they should check generated content against evidence-based medicine, expert consensus, and reliable sources.
The third competency, digital intelligence ethics and security, sets the baseline for responsible use. Medical education often involves clinical cases, patient information, academic writing, and learner assessment. Teachers therefore need to understand privacy protection, data de-identification, AI use disclosure, academic integrity, algorithmic bias, and risk control in AI-assisted decision-making. The consensus also emphasizes the importance of humanistic ethics, especially when AI is used in virtual patient interaction, case analysis, or clinical simulation.
The fourth competency, digital intelligence teaching integration, addresses how AI can be embedded into actual teaching. This includes the use of virtual patients, intelligent teaching assistants, AI-enhanced objective structured clinical examinations, adaptive learning systems, multimodal learning analytics, and AI-supported feedback. At the same time, teachers are expected to cultivate learners’ AI literacy, helping students critically evaluate AI outputs rather than passively accept them.
The fifth competency, digital intelligence research translation, is positioned as an advanced capacity for teaching leaders and expert faculty. It calls on educators to identify unmet needs in clinical teaching, translate them into research or engineering questions, and work with computer scientists, data scientists, and technology developers to build and evaluate intelligent teaching tools. The consensus makes clear that this is not a baseline requirement for all teachers, but a direction for faculty who lead educational innovation and translational work.
The framework was developed using a structured consensus process. The working group first conducted literature review and framework construction, then used brainstorming combined with a modified Delphi method for expert consultation. In the first round, 47 experts were invited and 45 valid responses were collected. In the second round, 35 experts participated and all questionnaires were valid. The expert authority coefficients were 0.824 and 0.813 in the two rounds, respectively, and tests of expert agreement reached statistical significance. Participants came from leading teaching hospitals, universities, and related institutions across China and Hong Kong, covering medical education, clinical teaching, teaching management, digital medicine, and AI technology.
The consensus also reflects current gaps in practice. The article notes that although Chinese medical teachers show growing awareness of generative AI, its use in teaching remains limited. In one cited survey of 456 critical care physicians in China, 64.70% reported using GenAI in clinical work, but only 33.10% used it in standardized residency training teaching activities. Among those who used GenAI in teaching, the most common uses were searching for teaching content and creating teaching materials. More than half of the physicians recognized ethical concerns, and 94.30% supported including ethics education in GenAI training.
These findings suggest that the main challenge is not whether medical teachers will encounter AI, but whether they are prepared to use it responsibly and effectively. Many teachers may already be experimenting with AI tools, but still lack systematic training in tool limitations, reliability verification, bias identification, privacy protection, and ethical governance.
For this reason, the consensus places strong emphasis on both capability and boundary-setting. It seeks to avoid two extremes: uncritical dependence on AI and complete rejection of AI. Instead, it proposes a structured approach in which teachers understand the technology, use it critically, manage risks, and integrate it into teaching only when it supports educational goals.
The consensus also connects teacher development with broader medical education reform. It can serve as a reference for medical schools, standardized residency training bases, and faculty development programs. Potential applications include setting training objectives, designing faculty development curricula, developing assessment tools, strengthening teacher training, and building digital teaching resources.
Looking ahead, the consortium and project working group plan to develop quantitative evaluation tools, including self-assessment scales and external review checklists, and to conduct multicenter validation studies. The article also calls for continued updates as generative AI technologies evolve, as well as stronger collaboration among educators, developers, administrators, and learners.
Generative AI is unlikely to replace medical teachers in a simple way. But it is changing the competencies teachers need. The central issue is no longer how many AI tools are introduced into medical education, but whether teachers can understand, supervise, and use these tools responsibly while continuing to guide clinical reasoning, professional judgment, and humanistic values.
The publication of this consensus provides a structured reference for medical teacher development in China’s digital transformation of medical education. It also offers a practical framework for thinking about what faculty training should look like as AI becomes increasingly embedded in clinical teaching and learning.
Article information
Title: Consensus on the Digital Intelligence Competency Framework for Medical Teachers by the China Consortium of Elite Teaching Hospitals
Journal: Chinese Journal of Medical Education Research
DOI: 10.3760/cma.j.cn116021-20260406-02299
Chinese Journal of Medical Education Research is a monthly peer-reviewed journal sponsored by the Chinese Medical Association and hosted by Chongqing Medical University, under the supervision of the China Association for Science and Technology.
Launched in 2002, the journal publishes research and practice-oriented studies on medical education, with a focus on teaching reform, clinical teaching, curriculum development, residency training, graduate education, nursing education, educational technology, and international medical education. It also features themed columns and special issues on emerging topics and institutional innovations in medical education.
The journal is recognized as a Source Journal for Chinese Scientific and Technical Papers and Citations and is included in the Chinese Science and Technology Core Journals. It is indexed in Wanfang Data, Index Copernicus, the WHO Western Pacific Region Index Medicus, and Ulrich’s Periodicals Directory.
Journal page: http://yxjyts.alljournals.ac.cn/homeNav?lang=zh
10.3760/cma.j.cn116021-20260406-02299
Literature review
People
Consensus on the Digital Intelligence Competency Framework for Medical Teachers by the China Consortium of Elite Teaching Hospitals
20-Jun-2026