Preamble

Review and Monograph
Research on Artificial Intelligence Ethics Literacy Among Medical School Teachers
WANG Zhenyu¹, GUO Pengfei¹*, HAO Wenhui²

¹School of Marxism, Chang'an University, Xi'an 710064, China
²Institute of Medical Information, Chinese Academy of Medical Sciences, Beijing 100020, China

Corresponding author: GUO Pengfei; E-mail: sjhdkahsaid@163.com

[Abstract] The application of artificial intelligence in medicine has become a crucial driving force for bridging the gap between basic research and clinical practice while reshaping diagnostic and therapeutic paradigms. As it transforms medical decision-making processes and doctor-patient interaction patterns, it also poses a paradigm challenge to traditional medical ethics education. In this context, the systematic construction of artificial intelligence ethics literacy among medical school teachers has emerged as a key issue in promoting the intelligent transformation of medical education. This study addresses the dilemmas of medical AI ethics education through an integrated approach of theoretical construction, mechanism analysis, and practical pathway design. From the intersection of philosophy of technology and medical ethics, we construct a literacy framework encompassing cognitive, competency, and value dimensions. We reveal the logic of ethics literacy generation through three-dimensional mechanisms of endogenous drive, exogenous synergy, and dynamic evolution, elucidating the interaction patterns among subject cognitive iteration, institutional contextual shaping, and risk adaptation. Finally, through the design of an integration framework for ethical cognition and technological practice, innovation in multi-modal practical training mechanisms driven by clinical scenarios, and institutional empowerment through ecological governance, we form a closed-loop collaborative cultivation system for AI ethics literacy among medical school teachers.

[Key words] Artificial intelligence; Medical school teachers; Ethical literacy; Teacher professional development; Medical education; Teacher quality; Collaborative cultivation

Funding: Shaanxi Provincial Social Science Fund Project (2023A029); 2022 Ministry of Education Special Research Project for Ideological and Political Theory Teachers in Higher Education (22JDSZK196)

Citation: WANG ZY, GUO PF, HAO WH. Research on artificial intelligence ethics literacy among medical school teachers[J]. Chinese General Practice, 2025. DOI: 10.12114/j.issn.1007-9572.2025.0187. [Epub ahead of print]. [www.chinagp.net]

1 Framework Construction for Artificial Intelligence Ethics Literacy Among Medical School Teachers

From ChatGPT to DeepSeek, generative artificial intelligence is undergoing a technological leap from general dialogue to deep vertical domain applications, triggering transformative waves across industries and continuously reshaping paradigms of production relations and social interaction. As General Secretary Xi Jinping pointed out: "We must strengthen the integration of artificial intelligence with efforts to safeguard and improve people's livelihoods" [1]. Currently, AI technology is accelerating the reconstruction of medical education models and clinical practice ecosystems. As dual-responsibility subjects for medical ethics education and technological innovation application, the cultivation of AI ethics literacy among medical school teachers has become a critical proposition for the innovative development of medical education systems in the era of intelligent healthcare.

Existing research on AI ethics literacy among medical school teachers has yielded numerous achievements. Regarding the connotation of AI ethics literacy, Zhao Yali et al. constructed an evaluation index system for intelligent general practitioners and weighted its indicators [2]. Fan Dingrong et al. developed a five-dimensional competency framework for "AI + nursing education" [3]. Regarding AI ethics risks, Lei Fang et al. [4] explored potential ethical risks in AI-based clinical decision support systems. Li Yiting et al. [5] investigated AI applications in nutrition management, while Yu Ruxia et al. [6] analyzed ethical risks that may arise from AI applications in clinical diagnosis and treatment. Regarding teacher competencies in medical schools, Huang Fumin et al. [7] analyzed the positive impacts of digital literacy on teaching quality, personalized learning, and teacher professional development. Yuan Na et al. [8] constructed a teaching efficacy scale for medical school teachers. While these findings provide important references for cultivating AI ethics literacy among medical teachers, there remains room for further optimization and enhancement. Particularly given the current triple dilemma facing medical AI ethics governance—ambiguous cognitive frameworks, unclear generation mechanisms, and unfocused practical pathways—existing studies either emphasize transplantation of universal AI ethics principles while neglecting medical professional characteristics, or remain limited to static factor analysis without revealing the dynamic evolution patterns of teacher AI literacy. More critically, they lack a complete research chain integrating theoretical construction, mechanism analysis, and pathway design from a systems perspective.

Therefore, it is essential to bridge the closed loop between theoretical construction and practical transformation in medical AI ethics education. Through mechanism deconstruction and pathway innovation, we must provide systematic solutions for cultivating medical educators with ethical consciousness and risk response capabilities, thereby facilitating the construction of a responsible new medical education ecosystem characterized by human-machine symbiosis.

This study employs a research methodology combining theoretical analysis and systematic construction from the interdisciplinary perspective of medical pedagogy and technology ethics. At the theoretical level, we use literature research and conceptual analysis to systematically review theoretical frameworks in bioethics and algorithmic ethics, revealing integration pathways between traditional medical ethics principles and AI ethics requirements through comparative studies. At the framework construction level, we adopt Korthagen's three-level teacher competency model as the theoretical foundation and combine it with the Delphi expert consultation method to build a competency framework encompassing technical cognition, ethical reflection, and educational practice dimensions. At the mechanism analysis level, we employ systems theory to analyze the interaction patterns of literacy generation from three dimensions: endogenous drive, exogenous synergy, and dynamic evolution. At the pathway design level, we use case studies and scenario simulation methods, combined with typical scenarios of medical AI applications, to propose clinically adaptable training programs.

The deep integration of AI technology in medicine has spawned new ethical issues that are both professional and technical, such as human-machine responsibility delineation and patient data privacy. As the educational responsibility subjects of medical ethics education, medical school teachers urgently need to construct an ethics literacy framework adapted to intelligent healthcare scenarios. Focusing on the theoretical foundation of integrating medical ethics and AI, we clarify the interactive relationship between their value logics and technical boundaries. We then analyze the core elements of AI ethics literacy for medical school teachers, establishing its connotation from cognitive, competency, and value dimensions. Finally, by constructing a multi-level, operational evaluation index system, we promote the transformation from theory to practice.

1.1 Overview of Artificial Intelligence Ethics Literacy for Medical School Teachers

Unlike traditional education, medical education is a social practice activity that purposefully, systematically, and organizationally cultivates medical talents in response to social needs. Medical school teachers are professional groups with dual identities as both medical practitioners and educators, representing composite roles integrating clinical medicine, educational theory, and humanistic literacy. Since medical education encompasses natural sciences, humanities, social sciences, and clinical disciplines, teachers must possess interdisciplinary knowledge integration capabilities. The World Federation for Medical Education (WFME) explicitly requires medical schools to establish teacher development mechanisms in its "WFME Global Standards for Quality Improvement in Basic Medical Education" [9], emphasizing that teachers must possess both medical professional competence and pedagogical literacy. The subsequent 2015 WFME standards [10] further strengthened this concept, requiring medical schools to "ensure positive interaction between medical research and teaching," implicitly demanding a dual-role design for teachers—requiring both medical expertise to ensure teaching accuracy and pedagogical skills to optimize teaching processes. Thus, medical school teachers must fulfill dual roles: both as disciplinary experts and as educators. Oleksandr Boychuk [11] argues that medical school teachers must achieve the professional cultivation and personality shaping of students (especially future healthcare workers) through "conscious and purposeful work," with their work possessing the duality of "organic unity of medical professionalism and social education." Based on this perspective, medical school teachers can be defined as professionals engaged in teaching, research, and clinical guidance in higher medical education institutions, with core responsibilities including imparting medical knowledge, cultivating clinical skills, guiding scientific research innovation, and shaping medical students' professional ethics.

In April 2021, during his inspection of Tsinghua University, General Secretary Xi Jinping emphasized the need to "advance the construction of new engineering, new medical science, new agricultural science, and new humanities, and accelerate the cultivation of urgently needed talents" [12]. The "Four New" initiative has become a landmark measure leading the reform and innovation of China's higher education. The proposal of "new medical science" aims to revolutionize talent cultivation models, not only training medical professionals but also cultivating medical scientists with research capabilities. Therefore, reform of the medical undergraduate teaching system in the new era is imperative, with teacher team construction being the primary task in this transformation. In this context, our understanding of medical school teachers needs to break through the traditional single-role positioning as "knowledge transmitters." Instead, we should view them as leaders in medical education reform, practitioners of interdisciplinary innovation, and promoters of the integration of research, clinical practice, and teaching. New medical science construction requires teachers to not only possess solid medical professional competence but also master emerging technologies such as artificial intelligence and big data [13] to meet the teaching demands of "medicine-engineering intersection" and "medicine-science fusion" [14].

Unlike AI ethics literacy in other fields, medical school teachers' AI ethics literacy has unique value concerns. First, the core dimensions of ethical consideration differ. Since medical AI applications directly concern life and health, teachers' ethics literacy must transcend general issues of data privacy or algorithmic fairness to address medical-specific ethical dilemmas. This requires ethical considerations to go beyond conventional technical ethics categories. Teachers need to guide students in deeply contemplating responsibility division in AI-assisted diagnosis, particularly the decision-making basis when algorithmic judgments conflict with clinical experience. Second, the integration methods of ethical norms differ. AI ethics in medical education cannot be separated from the existing medical ethics system. The primary value orientation of teacher education is to ensure students understand that AI introduction cannot weaken the fundamental principle of "patient benefit priority," thereby strengthening informed consent and minimizing harm in the context of new technologies. Finally, the practice orientation of ethics education differs. While AI ethics in other fields may focus on theoretical discussion, medical education must closely integrate with clinical scenarios, requiring teachers to cultivate students' ethical sensitivity through real cases. Moreover, the goal of medical AI ethics education is not only to avoid risks but also to maintain doctor-patient trust, ensuring that medical practice does not lose its humanistic essence in pursuit of efficiency.

Compared with clinical physicians, medical school teachers' AI ethics literacy demonstrates more distinctive group characteristics through their professional attributes. Medical school teachers need to focus on educational transmissive ethical decision-making abilities, while clinical physicians focus more on technical application ethics in clinical practice. Medical school teachers' AI ethics literacy manifests in: first, teaching transformation capability—the ability to translate abstract AI ethics principles into teachable cases that guide students to understand ethical conflicts behind technology; second, interdisciplinary critical thinking cultivation—revealing ethical paradoxes of AI in medical imaging diagnosis, gene editing, and other scenarios through curriculum design, such as algorithmic bias discrimination against vulnerable groups, which clinical physicians need to avoid in specific diagnosis and treatment; third, forward-looking ethics research—medical school teachers must track cutting-edge developments in AI ethics and continuously integrate them into teaching syllabus updates, while clinical physicians' ethical responsibilities focus more on technical usage compliance. This reflects the special mission of medical school teachers as the source of ethics education—their ethics literacy is not merely individual behavioral norms but a leverage point for shaping the overall AI ethics perspective of the future medical community.

1.2 Theoretical Interpretation Pathways for Artificial Intelligence Ethics Literacy in the Medical Field

The framework construction for AI ethics literacy in medicine must take the intersection and integration of bioethics and algorithmic ethics as its theoretical foundation, while accommodating both the particularity of medical practice and the universality of technical applications. Bioethics, as the core ethical framework in medicine, systematically explores moral and ethical issues in life sciences through the combination of science, philosophy, and ethics. It primarily focuses on interpreting the value, dignity, rights, and responsibilities of life from a medical perspective, emphasizing respect for and protection of life itself [15]. The basic framework of bioethics, comprising the principles of respect for autonomy, non-maleficence, beneficence, and justice proposed by American scholars Beauchamp and Childress, provides fundamental ethical constraints for AI applications in medical scenarios.

Specifically, the principle of respect for autonomy requires clinicians to inform patients in advance about the risks and uncertainties involved in using intelligent technology when adopting AI algorithm recommendations, respecting and maintaining patients' right to informed consent. The principle of non-maleficence relies on risk assessment mechanisms in AI clinical applications, maximizing the reduction of technical harms such as misdiagnosis and data leakage through algorithm auditing and clinical validation. The principle of beneficence focuses on the fundamental value orientation of "people-centeredness," with AI application research and development aiming to enhance diagnostic accuracy and accessibility as core objectives, creating optimal health benefits for patients through continuous algorithm performance optimization. The principle of justice emphasizes that AI medical resource allocation must avoid algorithmic bias, ensuring that patients of different genders, races, and socioeconomic statuses can equally access high-quality medical services. If AI models pursue only efficiency maximization, they may sacrifice the interests of vulnerable groups [16]. Principlism is the dominant methodology in AI ethics, representing the endogenous manifestation of translating AI ethics from moral principles to practice in medical contexts [17].

Algorithmic ethics, as a branch of technology ethics, focuses on ethical issues triggered by AI and big data technologies. Its core concern is how to embed moral principles into the core architecture and operational mechanisms of algorithms, enabling them to follow the value norms of human society during data processing and decision-making to ensure fairness, justice, and security in algorithm usage. In the medical field, algorithmic ethics primarily emphasizes transparency, interpretability, fairness, and accountability in technical applications. In clinical applications, it is necessary to disclose the training data sources, model architecture, and performance indicators of AI algorithms, understand algorithmic decision-making logic, ensure clinically verifiable algorithmic decision paths, and thereby avoid diagnostic risks caused by algorithmic "black box effects" [18].

The integration of bioethics and algorithmic ethics provides a solid theoretical foundation for interpreting AI applications in medicine and methodological guidance for addressing responsibility issues in ethical contexts. First, given the high-risk nature of medical decision-making, AI-empowered clinical diagnosis requires constructing a multi-level robust verification system that enhances the clinical reliability and safety of intelligent technology through adversarial sample testing and continuous learning mechanisms, continuously improving the anti-interference capability of medical AI systems. Second, addressing the essential attribute of "human care" in medicine requires further optimizing human-machine collaborative diagnosis and treatment models, fully maintaining the emotional connection in doctor-patient relationships by setting emotional interaction modules in intelligent consultation systems. Third, facing the dynamic complexity of medical problems, AI technology applications must be accompanied by adaptive ethical assessment frameworks that prevent diagnostic deviations caused by model hallucinations through multi-modal verification techniques [20].

1.3 Component Elements of Artificial Intelligence Ethics Literacy for Medical School Teachers

Teacher competency is an important research topic in pedagogy, with scholars developing multi-dimensional explanatory pathways such as the four-factor structure and the pyramid model for STEM teacher professional competency. AI ethics literacy for teachers represents a normative interpretation of specific dimensions in particular technical application fields. Fred Korthagen's three-level teacher competency model from Utrecht University in the Netherlands [21] provides a fitting explanatory pathway for interpreting the component elements of AI ethics literacy among medical school teachers. This model divides teacher behavior into three main levels: first, the basic technical operation level (based on tool usage experience and intuitive reactions); second, the ethical reflection integration level (forming embodied cognition through human-machine collaboration practice); and third, the systematic ethical decision-making level (involving multi-stakeholder interest balancing and value prioritization). These three levels form deep mapping with the practical requirements of AI ethics.

Based on this, using Korthagen's three-level model as a foundation, we can divide AI ethics literacy for medical school teachers into three dimensions. The technical cognition dimension focuses on teachers' foundational competencies as technology understanders, including two secondary indicators: AI technology understanding capability and data management capability. The former ensures teachers master the basic principles and limitations of medical AI, while the latter emphasizes the standardization of full-process medical data management. The ethical reflection dimension reflects teachers' professional role as ethical guides, with two secondary indicators: ethical principle application capability and bias identification and elimination capability, requiring teachers to both integrate traditional medical ethics with AI ethics and possess the ability to identify and eliminate algorithmic bias. The educational practice dimension highlights teachers' professional characteristics as educational implementers, containing three secondary indicators: ethical teaching design capability, student ethical literacy cultivation capability, and educational reflection and improvement capability, corresponding to three key educational links: curriculum development, student cultivation, and teaching optimization. This three-dimensional division follows the competency progression logic from technical cognition to ethical reflection to educational practice, fully embodying the professional characteristics of medical education and the dual mission of teachers, constructing a framework that meets both the common requirements of AI ethics and the distinctive features of medical education. The component framework of AI ethics literacy for medical school teachers is shown in Figure 1 [FIGURE:1].

2 Generation Mechanisms of Artificial Intelligence Ethics Literacy Among Medical School Teachers

The formation of teachers' AI ethics literacy is a multi-dimensional, dynamic, and complex process. From the theoretical perspectives of medical pedagogy and technology ethics, we can construct a three-dimensional collaborative evolution analysis framework based on the Embedded AI Ethics Education Framework and the Principle-Based Ethics Maturity Model [22]. First, at the individual cognition level, literacy generation stems from teachers' deep integration of medical ethical values and technical ethics requirements, manifested as a cognitive development trajectory from passive acceptance to active construction. This process relies on the three-dimensional competency structure of AI literacy—knowledge, skills, and attitudes—emphasizing teachers' understanding of AI principles, ability to identify ethical risks, and attitude toward responsible use [23]. Second, at the professional practice level, literacy development is shaped by the particularity of medical scenarios and the dual attributes of educational mission, strengthened through two-way feedback between clinical decision-making and teaching practice. This requires deeply integrating ethics education into medical curricula to form an embedded closed loop of curriculum development, teacher-student preparation, ethics integration, and practical feedback [24]. Finally, at the environmental adaptation level, literacy evolution manifests as continuous adaptation to technological transformation and ethics paradigm shifts, including both the identification and response to emerging ethical issues (such as algorithmic bias and data privacy) [25] and the creative transformation of traditional medical ethics principles. The logical structure of the generation mechanism for medical school teachers' AI ethics literacy is shown in Figure 2 [FIGURE:2].

2.1 Endogenous Drive Mechanism: Ethical Cognition and Technical Reflection of Medical School Teachers

The endogenous drive mechanism of medical school teachers' AI ethics literacy is rooted in their dual role positioning as both medical educators and industry practitioners, manifested as the systematic construction of teachers' ethical cognition systems and the conscious enhancement of technical reflection capabilities. In the AI era, medical school teachers must take medical ethics as the foundation, deeply integrate the technical principles of AI with the complexity of medical scenario practice, and systematically construct a cognitive framework guided by clear values, centered on deep risk identification, and aimed at practicing medical responsibility.

At the value orientation dimension, medical school teachers must deeply understand the extended connotations of core medical ethics values such as "life paramount," "patient rights priority," and "medical ethics as foundation" in the AI era. The UNESCO Recommendation on the Ethics of Artificial Intelligence passed in 2021 [26] proposes that AI development should always protect human rights, freedom, and dignity as core values. This forms a value alignment possibility with medical ethics values while requiring medical school teachers to fully clarify the ethical boundaries of intelligent technology tools in diagnosis, treatment, research, and teaching.

At the risk identification dimension, medical school teachers must systematically master the ethical risk landscape of AI applications in medical scenarios [27], understanding the forms and causes of risks such as data privacy leakage, algorithmic bias, and technological dependence. Particularly, they need to enhance their "questioning ability" regarding algorithms, actively "predicting" potential misdiagnoses or "hallucinations" that AI algorithmic models may produce in medical practice through proactive learning, continuously improving their ability to anticipate the limitations of algorithmic technology.

At the responsibility practice dimension, medical school teachers must enhance their subjective consciousness as technology leaders and become critical practitioners of AI technology ethics. Specifically, medical school teachers need to establish a two-way reflection mechanism in clinical teaching and research. On one hand, they must possess the ability to prudently analyze how AI optimizes diagnosis and treatment pathways, achieving precise traceability of every data link and conclusion generation. On the other hand, they must be highly vigilant against the erosion of medical humanistic spirit by technological alienation, avoiding the dissolution of student agency caused by technological alienation.

2.2 Exogenous Synergy Mechanism: Interaction Between Policy Norms and Medical Practice Contexts

The generation of AI ethics literacy among medical school teachers cannot be separated from systematic support from external policies and norms, while also requiring deep synergy with medical practice contexts to achieve competency enhancement. The exogenous synergy mechanism emphasizes constructing a full-chain teacher cultivation support system of "macro policy guidance—meso institutional design—micro behavioral norms" through institutional constraints, contextualized adaptation, and interdisciplinary collaboration.

At the macro policy guidance dimension, focusing on the "New Generation AI Ethics Norms" [28] and the "Opinions on Accelerating Educational Digitalization" [29], we must further strengthen top-level national design for medical AI ethics, clarifying the AI ethics assessment capabilities that teachers must possess, and providing policy basis and directional guidance for cultivating medical school teachers' AI ethics literacy.

At the meso institutional design dimension, we must further implement teacher cultivation systems based on industry standards such as "Teacher Digital Literacy" [30] and local programs like the "Shanghai Medical AI Work Plan (2025-2027)" [31], while establishing interdisciplinary ethics review mechanisms within the industry, integrating AI ethics into teacher assessment systems, and providing institutionalized training and resource support for teachers to ensure competency enhancement is closely integrated with clinical practice.

At the micro behavioral norms dimension, we must pay attention to the particularity of medical school education, especially focusing on each institution's professional strengths and medical talent cultivation directions, promoting the establishment of precise school-specific implementation plans for teachers' AI literacy [32], and developing school-based AI application ethics operation manuals that refine operational standards for data collection, algorithm training, and clinical deployment.

Furthermore, medical schools should conduct training that embeds relevant AI technology ethics norms into real medical scenarios, improving teachers' clinical practice capabilities through contextualized ethics training. Schools can establish ethics workshops for medical scenario applications, where teachers participate in attribution analysis of real cases such as AI misdiagnosis disputes and patient data leakage incidents, extract and summarize ethical risk prevention and control strategies, and transform reflection outcomes into teaching resources, strengthening their practical transformation ability of technical ethics cognition. Schools can also establish development portfolios for teachers' AI ethics literacy, recording dynamic data on teachers' participation in technical ethics training, technical review projects, and academic outputs, while introducing peer review and patient feedback mechanisms to form a feedback closed loop from learning to practice to evaluation and improvement, promoting continuous self-assessment and knowledge updating among medical school teachers, and driving teachers to internalize technical ethics norms as professional instincts, enabling them to always adhere to the bottom line of medical ethics in technology applications.

2.3 Dynamic Evolution Mechanism: Continuous Adaptation to Technical Risks and Medical Ethics Conflicts

The generation of AI ethics literacy among medical school teachers is a dynamic evolution process that requires constructing a rapid response system adapted to fast technological iteration and escalating ethics challenges. As AI application scenarios continuously enrich and expand, the ethical dilemmas they trigger also show dynamic change trends. Teachers' ability to grasp evolving ethical focuses from the technological frontier has become an important criterion for measuring their literacy and even competency.

First, medical schools must establish dynamic monitoring and tiered response mechanisms for technical ethics risks, providing real-time updated ethical decision-support tools and dynamic tiered training resources for teachers' AI ethics literacy. By fully utilizing industry resources and collaborating with affiliated hospitals and enterprises, they should develop ethical risk matrices for medical AI, real-time tracking and analysis of the latest ethical controversies, policy and regulatory changes, and academic research progress in medical AI applications both domestically and internationally. Simultaneously, they should conduct risk assessment and tiered ranking of medical AI tools from dimensions such as technology maturity, data sensitivity, and clinical impact scope. For example, autonomous surgical robots should be classified as high-risk technology, requiring restricted usage scenarios and strengthened pre-operative ethics review, while AI-assisted imaging diagnosis should be classified as medium-risk, requiring regular verification of algorithmic fairness. This dynamic risk management model can continuously expose teachers to new ethics challenges, continuously updating their ethical cognition frameworks and decision-making abilities, thereby transforming theoretical norms into practical wisdom in specific contexts and achieving spiral upgrading of ethics literacy.

Second, the systematic enhancement of medical school teachers' AI ethics literacy requires continuous adaptation to medical ethics conflicts. With the emergence of new ethical dilemmas such as conflicts between algorithmic decision-making and clinical experience, and tensions between patient privacy protection and data sharing, higher requirements are placed on teachers' ethical judgment abilities. As a long-term organizational response strategy, interdisciplinary ethics case seminars can be organized to guide teachers in analyzing ethical dilemmas in real clinical scenarios, while encouraging teachers to participate in ethics review work for medical AI projects, deepening their understanding of technical ethics conflicts in practice. This continuous adaptation process enables teachers to dynamically grasp ethics changes brought by technological development, transforming abstract ethical principles into concrete coping strategies, thereby forming clinically adaptable ethics literacy.

3 Cultivation Pathways for Artificial Intelligence Ethics Literacy Among Medical School Teachers

As AI technology reshapes the knowledge graph and practice fields of medical education, medical school teachers' ethical roles are shifting from traditional transmitters of ethical principles to value guides for human-machine collaborative decision-making. However, existing teacher cultivation systems in medical schools still suffer from problems such as disconnection between ethical cognition and technical application, insufficient clinical contextual adaptation, and institutional support gaps, making it difficult to address complex ethical risks such as AI diagnosis algorithm bias and data privacy leakage. Based on this, we need to resolve the structural dilemmas in developing medical school teachers' AI ethics literacy during the intelligent transformation of medical education through framework design, pathway innovation, and institutional optimization, providing systematic solutions for constructing a responsible medical AI education ecosystem.

3.1 Integration of Ethical Cognition Construction and Technical Practice: Cultivation Framework Design for Medical School Teachers' AI Ethics Literacy

The cultivation of medical school teachers' AI ethics literacy must center on systematic knowledge construction and integration with technical practice capabilities, building a progressive cultivation framework of theoretical foundation, skill strengthening, and value internalization. At the theoretical foundation level, teacher cultivation should focus on constructing interdisciplinary knowledge systems, integrating knowledge from medical ethics, data governance, and AI technology principles. The medical ethics module focuses on analyzing programmatic documents such as the Declaration of Geneva [33] and the New Generation AI Ethics Norms [25], strengthening teachers' deep understanding of "respect for patient autonomy," "non-maleficence principle," and "fairness," and clarifying the reshaping logic of doctor-patient trust relationships by AI technology through comparing differences in rights and responsibilities between traditional medical ethics and the AI era. The data governance module focuses on the specific application of regulations such as the Personal Information Protection Law of the People's Republic of China [34] and the Biosecurity Law of the People's Republic of China [35] in medical scenarios, requiring teachers to master operational norms such as de-identification, encrypted transmission, and compliant sharing of patient biometric data, and enhancing teachers' risk prediction capabilities through case analysis of ethical controversies such as cross-border use of genetic data. The AI technology module emphasizes the correlation teaching between AI technology principles and ethical risks, such as explaining how the "black box" characteristics of deep learning models lead to non-traceable diagnostic results, or how the "hallucination" phenomenon of generative AI may trigger misdiagnosis risks.

Innovation is also needed in methods and approaches for cultivating medical school teachers' AI ethics literacy, constructing a three-dimensional cultivation model combining blended learning, immersive practical training, and reflective seminars. Blended learning relies on platforms such as MOOCs and Xuexitong to provide basic theoretical courses for teachers' autonomous knowledge accumulation. Immersive practical training conducts ethics review simulations through medical simulation centers. For example, regarding deployment applications for AI-assisted diagnosis systems, medical school teachers must simulate the role of ethics committees, reviewing data source compliance, algorithm transparency reports, and emergency plans, and writing evaluation opinions. The reflective seminar component introduces an "ethics roundtable" mechanism, inviting clinicians, legal consultants, and patient representatives to jointly discuss boundaries of AI technology applications, prompting medical school teachers to deepen ethical cognition through multi-perspective collisions.

3.2 Clinical Scenario Embedding and Multi-Modal Practical Training Synergy: Competency Advancement Pathways for Medical School Teachers' AI Ethics Literacy

The competency advancement pathways for medical school teachers' AI ethics literacy need to be based on the national new medical science development strategy, conform to the trend of medical intelligent transformation, and construct a clinically embedded cultivation system with Chinese characteristics. This cultivation system must promote medical model reforms based on general education, strengthen the combination of basic and clinical medical education, integration of research training and medical practice, and integration of professional humanistic literacy and medical education [36], achieving systematic enhancement of teachers' AI ethics literacy through coordinated multi-modal practical training.

At the basic competency cultivation stage, we must integrate medical education characteristics to implement relevant policy requirements from the New Generation AI Ethics Norms [25], organically combining ethics awareness cultivation with teachers' teaching competency enhancement. Organize teachers to participate in the full-process management of outpatient AI-assisted diagnosis systems in rotation, including data collection, algorithm application, and result verification, cultivating their cognitive abilities regarding basic ethical issues such as medical data privacy protection and algorithmic fairness. Simultaneously, conduct standardized simulation training based on real cases to continuously expose teachers to real doctor-patient communication scenarios in AI consultations, helping teachers deepen their understanding of technical ethics in clinical practice. This stage primarily cultivates teachers' general understanding of AI medical scenarios, ensuring teacher competency development keeps pace with cutting-edge industry development.

At the professional competency deepening stage, we must fully leverage the demonstration and leading role of regional high-level hospitals, utilizing platform resources from top-tier affiliated hospitals to echelon-build demonstration bases for medical AI ethics practical training. Through hospital-school collaboration, construct high-fidelity clinical training environments such as surgical robots and intelligent imaging diagnosis, and conduct immersive, interactive multi-modal practical training. This stage focuses on cultivating teachers' ethical decision-making abilities in complex medical scenarios. Through typical situations such as early warning handling of intelligent monitoring systems in intensive care units and resource allocation in AI-assisted emergency triage, teachers can transform clinical participation experiences into classroom teaching cases, forming a teaching case repository based on practical resources. Simultaneously, we must further promote the establishment of interdisciplinary and cross-institutional collaborative cultivation mechanisms, integrating expert resources from multiple fields such as clinical medicine, AI, and ethics to provide comprehensive professional support for teacher development.

At the comprehensive competency enhancement stage, we must focus on cultivating medical education leaders with international perspectives. Select outstanding frontline teachers to participate in national medical AI ethics standard formulation and industry norm revision, bringing real medical education contexts into policy formulation frameworks. Simultaneously, establish transformation mechanisms from clinical practice to teaching innovation, converting cutting-edge technical scenarios such as surgical robot applications and gene editing therapy into teaching cases and curriculum resources. Build a "government-industry-academia-research-application" collaborative innovation platform to promote deep cooperation among medical institutions, universities, and enterprises, jointly exploring new models of medical AI ethics education. This stage's cultivation must be based on the developmental needs of medical education modernization, cultivating high-quality medical education talents for building a global community of health for all. Through this hierarchical, multi-measure cultivation pathway, we ultimately achieve teachers' strategic transformation from technology users to ethics leaders, providing strong talent support for Healthy China construction.

3.3 Institutional Empowerment and Ecological Governance: Long-Term Guarantee Mechanism for Medical School Teachers' AI Ethics Literacy

The sustainable development of medical school teachers' AI ethics literacy requires constructing a systematic, multi-level institutional guarantee system. At the policy guidance level, we must promptly promote collaborative formulation of cultivation guidelines for medical school teachers' AI ethics literacy by medical industry management departments, education management departments, and science and technology management departments, explicitly incorporating AI ethics literacy into the core indicators of teacher professional development.

At the practical implementation level, first, we can establish a mandatory continuing education credit system, requiring medical school teachers to complete fixed credits of AI ethics thematic training annually. Additionally, we can add an evaluation dimension of "AI ethics application capability" in the professional title review system, incorporating participation in medical AI ethics review and development of relevant teaching cases into review indicators. Furthermore, we can establish special development channels, providing key cultivation support for teachers with outstanding performance in AI ethics research and teaching.

In terms of evaluation and incentives, we must further improve the whole-process, multi-dimensional evaluation system that incorporates AI ethics indicators for medical school teachers. First, we should scientifically develop standardized evaluation tools, including AI ethics knowledge question banks (covering core knowledge points such as data privacy and algorithmic bias) and scenario simulation assessment systems (such as AI-assisted diagnosis dispute handling simulations). Second, we can gradually promote the implementation of clinical practice competency assessment, giving frontline teachers more opportunities to participate in real cases and incorporating these into relevant evaluation criteria. Additionally, we can establish a contribution point system, providing quantitative evaluation for achievements such as participating in industry standard formulation, publishing teaching cases, and exploring AI-empowered teaching method optimization. Finally, we can link evaluation results directly to teacher development, using them as references for position appointment and professional title promotion.

In terms of resource synergy, we must further build a "government-industry-academia-research-application" integrated support network for medical school teachers. At the government level, we should further expand the national medical AI ethics data platform, integrating resources such as typical ethics cases and policies and regulations to provide authoritative references for teachers. At the industry level, we should rely on national medical centers to establish several regional teacher development bases, equipped with facilities such as intelligent consultation simulation systems and surgical robot training platforms, providing rotation training for medical school teachers nationwide. At the enterprise level, we should promote leading medical AI enterprises and medical schools to jointly build joint laboratories, developing specialized ethics practical training systems. At the academic level, we should establish a medical AI ethics education alliance, regularly holding teaching competitions and conducting international exchanges. Through this three-dimensional guarantee mechanism, we ultimately achieve the standardization, normalization, and sustainable development of medical school teachers' AI ethics literacy cultivation.

Conclusion

This study systematically constructs a theoretical framework and practical pathways for AI ethics literacy among medical school teachers, revealing its unique value as a key element in the fusion of medical education and technological innovation. The research finds that the development of medical school teachers' ethics literacy exhibits competency advancement characteristics from technical cognition to ethical judgment to educational practice. This process requires both teachers' deepening understanding of technical ethics in medical scenarios and systematic support from the institutional environment. The clinical scenario embedding and multi-modal practical training synergy cultivation pathways proposed in this study effectively promote the transformation from theoretical cognition to practical capability by converting typical medical cases into teaching resources. However, with the rapid development of new technologies such as generative AI, medical education faces new challenges such as algorithmic transparency and data privacy protection, requiring teacher ethics literacy cultivation to maintain dynamic adaptability. Future research should, based on the existing theoretical framework, further explore differentiated cultivation strategies across various medical scenarios, improve evaluation index systems, and strengthen international comparative studies to build a more inclusive and forward-looking medical AI ethics education system, providing continuous support for cultivating medical education talents in the new era.

References

[1] XI Jinping emphasized during the ninth collective study session of the Political Bureau of the CPC Central Committee: strengthen leadership, make good plans, clarify tasks, and consolidate foundations to promote the healthy development of China's new generation of artificial intelligence [N]. People's Daily, 2018-11-01(01).

[2] ZHAO Yali, LU Xiaoqin, LIU Jue, et al. Construction of an evaluation index system for intelligent general practitioners [J]. Chinese General Practice, 2025, 28(22): 2705-2711. DOI: 10.12114/j.issn.1007-9572.2025.0020.

[3] FAN Dingrong, TANG Yinghong, SHEN Ao, et al. Analysis of core competencies of nursing teachers under the "AI + nursing education" background [J]. China Continuing Medical Education, 2025, 17(8): 150-154. DOI: 10.3969/j.issn.1674-9308.2025.08.032.

[4] LEI Fang, DU Liang, DONG Min, et al. Transparent reporting of early clinical evaluation of artificial intelligence-based clinical decision support systems [J]. Chinese General Practice, 2024, 27(10): 1267-1270. DOI: 10.12114/j.issn.1007-9572.2023.0668.

[5] LI Yiting, TU Wenjing, YIN Tingting, et al. Application of artificial intelligence in nutrition management of patients with inflammatory bowel disease: a scoping review [J]. Chinese General Practice, 2025, 28(14): 1709-1716. DOI: 10.12114/j.issn.1007-9572.2024.0276.

[6] YU Ruxia, JIANG Jing, WANG Qiucheng, et al. Research hotspots and frontier trend analysis of artificial intelligence in clinical diagnosis and treatment of Alzheimer's disease [J]. Chinese General Practice, 2024, 27(26): 3218-3226. DOI: 10.12114/j.issn.1007-9572.2023.0704.

[7] HUANG Fumin, YAN Hongyu, JIA Qiannan, et al. Data-driven education transformation can enhance digital literacy of medical school teachers [J]. Basic & Clinical Medicine, 2024, 44(6): 897-900. DOI: 10.16352/j.issn.1001-6325.2024.06.0897.

[8] YUAN Na, JIAO Nan. Revision and reliability and validity test of the Chinese version of the teaching efficacy scale for medical school teachers [J]. TCM Education, 2023, 42(5): 39-42. DOI: 10.3969/j.issn.1003-305X.2023.05.158.

[9] World Federation for Medical Education, Secretariat of Clinical Medicine Professional Accreditation Working Committee of Ministry of Education of China. WFME global standards for quality improvement in basic medical education (2012 revised edition) [J]. Chinese Journal of Medical Education, 2014, 34(3): 321-334. DOI: 10.3760/cma.j.issn.1673-677X.2014.03.001.

[10] World Federation for Medical Education, Secretariat of Clinical Medicine Professional Accreditation Working Committee of Ministry of Education of China. WFME global standards for quality improvement in basic medical education (2015 revised edition) [J]. Chinese Journal of Medical Education, 2019, 39(6): 401-419. DOI: 10.3760/cma.j.issn.1673-677X.2019.06.001.

[11] BOYCHUK A. Сучасні психолого-педагогічні вимоги до викладача вищого медичного навчального закладу [J]. Educological discourse, 2018: 158-170.

[12] Adhere to the goal and direction of building world-class universities with Chinese characteristics and contribute to serving national prosperity, national rejuvenation, and people's happiness [N]. People's Daily, 2021-04-20(01).

[13] WANG Licong, WANG Yu, WANG Bo, et al. Research on artificial intelligence empowering faculty construction in medical schools [J]. China Medical Education Technology, 2025, 39(4): 450-454. DOI: 10.13566/j.cnki.cmet.cn61-1317/g4.202504005.

[14] HE Wei, ZHENG Yudong, HE Zhiwei. Preliminary exploration of the "industry-education-medicine trinity" training model for professional degree postgraduates in the interdisciplinary field of medicine and engineering [J]. Journal of University of Science and Technology Beijing (Social Sciences Edition), 2025, 41(3): 30-36. DOI: 10.19979/j.cnki.issn10082689.2024040057.

[15] LI Jianhui. Human dignity and bioethics [J]. Medicine and Philosophy (Humanistic & Social Medicine Edition), 2008, 29(11): 34-36, 42. DOI: CNKI:SUN:YXZX.0.2008-11-014.

[16] RAJKOMAR A, HARDT M, HOWELL M D, et al. Ensuring fairness in machine learning to advance health equity [J]. Ann Intern Med, 2018, 169(12): 866-872. DOI: 10.7326/M18-1990.

[17] LI Xueyao. Interdisciplinary construction of artificial intelligence ethics system: thinking from the perspective of complex adaptive systems [J]. Zhejiang Academic Journal, 2024(6): 48-56. DOI: 10.16235/j.cnki.33-1005/c.2024.06.001.

[18] OXHOLM C, CHRISTENSEN A S, NIELSEN A S. The ethics of algorithms in healthcare [J]. Camb Q Healthc Ethics, 2022, 31(1): 119-130. DOI: 10.1017/S0963180121000864.

[19] WANG Zhongyan, AN Na. Return and reconstruction of medical humanistic spirit [J]. Chinese Medical Ethics, 2008, 21(4): 145-146. DOI: 10.3969/j.issn.1001-8565.2008.04.060.

[20] LI Xu, ZHU Rui, CHEN Xiaolei, et al. A survey on hallucination of vision language large models: causes, evaluation, and governance [J/OL]. Journal of Computer Research and Development, 1-24. [2025-05-21]. http://kns.cnki.net/kcms/detail/11.1777.TP.20250506.1509.006.html. DOI: 10.7544/issn1000-1239.20240506.

[21] KORTHAGEN F A J. Making teacher education relevant for practice: the pedagogy of realistic teacher education [J]. Orbis Scholae, 2018, 12(2): 125-140. DOI: 10.14712/23363177.2018.99.

[22] WEIDENER L, FISCHER M. Proposing a principle-based approach for teaching AI ethics in medical education [J]. JMIR Med Educ, 2024, 10(1): e55368. DOI: 10.2196/55368.

[23] HU Wei. How can artificial intelligence empower teacher development: components and generation pathways of teacher AI literacy [J]. Journal of Teacher Education, 2024, 11(2): 39-47. DOI: 10.13718/j.cnki.jsjy.2024.02.005.

[24] QUINN T P, COGHLAN S. Readying medical students for medical AI: The need to embed AI ethics education [EB/OL]. (2021-09-08) [2025-06-24]. https://arxiv.org/abs/2109.02866. DOI: 10.48550/arXiv.2109.02866.

[25] ALONSORODRIGUEZ A M. Towards an ethical framework for artificial intelligence in education [J]. Teoria de la Educacion, 2024, 36(2): 79-98. DOI: 10.14201/teri.31821.

[26] ZHU Liyu, HU Xiaofan. UNESCO's Recommendation on the Ethics of Artificial Intelligence and its Chinese contributions from the perspective of human rights protection [J]. Human Rights, 2022(4): 47-64.

[27] TONG Ruipeng, HU Xiangyang, HE Jiehua, et al. Risk zoo (I): composition and representation of risk mapping based on metaphor theory [J]. China Safety Science Journal, 2023, 33(6): 27-34. DOI: 10.16265/j.cnki.issn1003-3033.2023.06.004.

[28] Ministry of Science and Technology of the People's Republic of China. Release of the New Generation AI Ethics Norms [EB/OL]. (2021-09-26) [2025-07-30]. https://www.most.gov.cn/kjbgz/202109/t20210926_177063.html.

[29] Ministry of Education of the People's Republic of China. Opinions of the Ministry of Education and nine other departments on accelerating educational digitalization [EB/OL]. (2025-04-15) [2025-07-30]. http://www.moe.gov.cn/srcsite/A01/s7048/202504/t20250416_1187476.html.

[30] Ministry of Education of the People's Republic of China. Notice of the Ministry of Education on issuing the "Teacher Digital Literacy" education industry standard [EB/OL]. (2022-12-02) [2025-07-30]. http://www.moe.gov.cn/srcsite/A16/s3342/202302/t20230214_1044634.html.

[31] Shanghai Municipal People's Government. Notice of the General Office of Shanghai Municipal People's Government on issuing the "Shanghai Medical AI Development Work Plan (2025-2027)" [EB/OL]. (2024-12-24) [2025-07-30]. https://www.shanghai.gov.cn/nw12344/20241224/0f60c01551784720899dbb911e9d5f08.html.

[32] KITAGAWA F, SÁNCHEZ BARRIOLUENGO M, UYARRA E. Third mission as institutional strategies: between isomorphic forces and heterogeneous pathways [J]. Sci Public Policy, 2016: scw015. DOI: 10.1093/scipol/scw015.

[33] World Medical Association. WMA Declaration of Geneva [EB/OL]. (2024-05-31) [2025-07-30]. https://www.wma.net/policies-post/wma-declaration-of-geneva/.

[34] Ministry of Education of the People's Republic of China. Personal Information Protection Law of the People's Republic of China [EB/OL]. (2021-08-20) [2025-07-30]. https://www.gov.cn/xinwen/2021-08/20/content_5632486.htm.

[35] Ministry of Ecology and Environment of the People's Republic of China. Biosecurity Law of the People's Republic of China [EB/OL]. (2020-10-17) [2025-07-30]. https://www.mee.gov.cn/ywgz/fgbz/fl/202303/t20230314_1019536.shtml.

[36] PENG Shutao. The concept and action of "new medical science" [J]. Journal of Shanghai Jiao Tong University (Philosophy and Social Sciences), 2020, 28(5): 145-152. DOI: 10.13806/j.cnki.issn1008-7095.2020.05.014.

Author Contributions: WANG Zhenyu proposed the research idea and designed the research plan; GUO Pengfei was responsible for research implementation; HAO Wenhui was responsible for data collection, acquisition, and analysis; WANG Zhenyu, GUO Pengfei, and HAO Wenhui jointly participated in discussion, revision, and improvement at all stages of the paper; WANG Zhenyu was responsible for final version revision and is accountable for the paper.

Conflict of Interest: The authors have no conflicts of interest to declare.

ORCID IDs:
WANG Zhenyu https://orcid.org/0009-0001-6945-0697
GUO Pengfei https://orcid.org/0009-0005-6880-8801

Received: 2025-06-30; Revised: 2025-07-31

Editor: KANG Yanhui