The core of educational digital transformation lies in the digital readiness of people. As the primary facilitators of educational activities, the degree to which teachers' instructional capabilities are intelligently adapted directly determines the effectiveness of technology-enabled education. Generative AI has reshaped the entire process of educational content production, teaching implementation, and learning outcome evaluation, placing new demands on teachers' AI literacy, instructional innovation capacity, and human-AI collaboration skills. However, existing research indicates that the practical AI application skills among primary and secondary school teachers in China remain relatively low, alongside issues such as disconnects between training content and teaching practice, homogenization of training models, and formalism in effect evaluation. In this context, constructing a training system for teachers' instructional capability that meets the demands of the intelligent era is both a practical response to the educational digitalization strategy and an inevitable choice for fostering a paradigm shift in teacher professional development. Based on the value dimension of AI empowerment, this study systematically explores the construction logic and implementation pathways of the training system, intending to provide a theoretical reference for regional and inter-school teacher training practices.

1. The Logical Core of AI-Enabled Enhancement of Teachers' Instructional Capability

1.1 The Coupling Relationship between AI and Teachers' Instructional Capability

The coupling between AI and teachers' instructional capability is essentially a bidirectional adaptation between technological logic and educational logic. From the perspective of technological logic, generative AI possesses core functions such as content generation, data analysis, and personalized adaptation, enabling it to take over repetitive, standardized teaching tasks (e.g., drafting lesson plans, grading assignments, compiling learning data statistics) from teachers. This liberation from lower-order tasks allows teachers to focus on higher-order educational aspects such as instructional design, emotional support, and cognitive guidance. This technological characteristic necessitates an upgrade in the system of teachers' instructional capabilities from "technology application" to "technology integration," specifically manifested as: the ability to select and apply AI tools appropriately based on teaching scenarios; the ability for human-AI collaborative instructional design, integrating AI technology into all phases of design including learning objectives, teaching processes, and evaluation; and the capacity for data-driven teaching reflection, optimizing teaching strategies based on learning analytics generated by intelligent platforms.

From the perspective of educational logic, the core of teachers' instructional capability remains their educational competence—their ability to foster student development. The integration of AI is not intended to replace teachers but to reinforce their central role through technological empowerment. The coupling of the two must adhere to the principle of "teaching determined by learning, with AI as an aid," avoiding a tendency towards technological instrumentalism and ensuring that AI serves the educational goal of holistic student development. This coupling relationship dictates that training for enhancing teachers' instructional capability cannot focus solely on technical operations but must integrate and enhance multi-dimensional literacies encompassing tool application, instructional design, and ethical awareness.

1.2 The Value Dimensions of AI-Enabled Enhancement of Teachers' Instructional Capability

AI empowerment for enhancing teachers' instructional capability encompasses both instrumental rationality and value rationality. At the level of instrumental rationality, technological empowerment directly enhances the efficiency and precision of teaching activities: intelligent lesson preparation platforms can rapidly generate personalized lesson plans and resources, reducing teachers' preparation time; intelligent learning analytics systems can accurately identify students' knowledge gaps, enabling targeted teaching interventions; intelligent evaluation tools can automate the collection and analysis of process-oriented assessment data, enhancing the comprehensiveness and objectivity of evaluation. These benefits pertain to the "efficiency enhancement" dimension of teachers' instructional capability, constituting the foundational value of technological empowerment.

At the level of value rationality, technological empowerment promotes a return of teachers' instructional capability to the "essence of education." With AI handling lower-order teaching tasks, teachers' core work shifts towards building supportive teacher-student relationships, designing profound learning experiences, and guiding students' value formation. This requires teachers to strengthen core competencies such as humanistic literacy, curriculum creation capacity, and interdisciplinary teaching ability. Simultaneously, the application of AI also demands enhanced ethical cognitive abilities from teachers, including the capacity to discern the accuracy of AI-generated content, safeguard student data privacy, and guide students in the appropriate use of AI. These constitute the advanced value dimension of technological empowerment.

2. Framework for Constructing the AI-Empowered Training System for Enhancing Teachers' Instructional Capability

2.1 Fundamental Principles for Constructing the Training System

2.1.1 Demand-Orientation Principle

The construction of the training system must be based on the stratified and individualized needs of the teacher population. Methods such as big data analysis, questionnaires, and interviews should be employed to accurately identify the competency gaps and developmental demands of teachers across different educational stages, years of experience, and subject areas. For instance, school-level administrators need enhanced educational leadership for the AI era, frontline teachers require strengthened contextualized technology application skills, and teacher trainers need improved guidance capabilities for AI application. This principle ensures that training content aligns with teachers' actual teaching contexts, avoiding a homogenized "one-size-fits-all" approach.

2.1.2 Principle of Human-Machine Collaboration

The training system must embed the core concept of "human-machine collaboration," clearly defining AI's supportive role in teaching while avoiding technology-centric biases. Beyond operational skills for intelligent tools, the curriculum should encompass methodologies for human-machine collaborative instructional design. This guides teachers to comprehend technology's applicable boundaries, master a "teacher-led, AI-assisted" pedagogical framework, and preserve education's humanistic nature alongside teachers' principal status.

2.1.3 Dynamic Adaptation Principle

Given the rapid iteration of AI technologies and the dynamic nature of teaching scenario demands, the training system must possess dynamic adjustment capabilities. Training content needs timely updates following technological advancements; for example, new features and application scenarios of generative AI should be quickly incorporated into the training scope. Training methods must adapt to teachers' learning habits, blending online intelligent learning platforms with offline focused workshops to achieve ubiquitous and personalized learning support.

2.1.4 Literacy Integration Principle

The training system must integrate technological literacy and pedagogical literacy. Building on the TPACK (Technological Pedagogical Content Knowledge) framework, AI literacy should be embedded within the integrated knowledge base of content, pedagogy, and technology, preventing a disconnect between technical training and teaching practice. The training objective targets teachers' integrated capability in "AI + subject teaching," rather than isolated technical operational skills.

2.2 Core Components of the Training System

2.2.1 Training Objectives

Training objectives should be set hierarchically: For school-level leaders, the goal is to cultivate their educational planning capability and leadership in the AI era, enabling them to formulate school-level AI-in-education application plans. For subject-specialist master teachers, the goal is to foster their innovation and demonstration capacity in AI application, enabling them to develop "AI + subject" teaching cases and provide mentorship. For general frontline teachers, the goal is to develop their foundational application skills for AI tools and human-AI collaborative instructional design ability, enabling them to integrate technology into daily teaching. For teacher trainers and researchers, the goal is to cultivate their guidance and evaluation capabilities for AI application, enabling them to provide precise training support to frontline teachers.

2.2.2 Training Content

Training content is structured into three dimensions: the foundational layer, the integration layer, and the advanced layer. The foundational layer focuses on basic AI knowledge and tool application, including the operation of mainstream intelligent teaching tools, data privacy, and ethical norms. The integration layer emphasizes human-AI collaborative instructional design, encompassing AI-based learning analytics, personalized teaching plan design, and the application of intelligent evaluation tools. The advanced layer concentrates on teaching innovation and research capabilities, including the development of AI teaching application cases, data analysis and reflection on teaching effectiveness, and the construction of interdisciplinary intelligent teaching models.

2.2.3 Training Providers

A quaternary collaborative training provider system involving "Universities + Training/Research Institutions + EdTech Companies + Schools" is constructed. Universities are responsible for theoretical guidance and disseminating cutting-edge research findings; training/research institutions are responsible for the design and implementation of training programs; EdTech companies are responsible for providing technical tools and operational guidance; schools are responsible for offering teaching scenarios and practical validation. The four entities fulfill their respective roles, leveraging their complementary strengths to form a collaborative mechanism characterized by "joint needs assessment — co-development of curricula — collaborative problem-solving — sharing of outcomes."

2.2.4 Training Platforms

Intelligent learning platforms serve as the core carrier, integrating online and offline training resources. The online platform should feature functions such as course learning, tool practice, case sharing, and learning diagnosis, supporting teachers' ubiquitous and personalized learning. Offline platforms include centralized workshops, seminars, school-based research activities, etc., focusing on solving authentic problems in teaching practice to achieve deep integration of theory and practice.

3. Implementation Pathways for the AI-Empowered Training System to Enhance Teachers' Instructional Capability

3.1 Precise Needs Analysis Prior to Implementation

Precise needs analysis is the prerequisite for implementing the training system and requires leveraging big data technology to construct teacher competency profiles. First, establish an assessment indicator system for teachers' AI competency, designing evaluation dimensions from the three aspects of knowledge, skills, and attitude, covering core indicators such as AI awareness, tool application, instructional design, and ethical cognition. Second, utilize intelligent assessment platforms to conduct quantitative evaluations of teachers' competencies, combined with qualitative methods like interviews and classroom observations, to identify specific ability gaps and developmental needs. Finally, based on the assessment results, categorize teachers into different tiers, providing a basis for stratified training design and ensuring that training content accurately matches teachers' starting points and developmental requirements.

3.2 Stratified Training Design During Implementation

Stratified training design must align with the developmental objectives of teachers at different levels, employing differentiated training methods and content: For the leader tier, combine expert guidance, case study discussions, and plan development activities, focusing on cultivating educational management and strategic planning capabilities for the AI era. For the master teacher tier, employ methods such as workshops and project-based learning, focusing on fostering innovation and demonstration capabilities in AI teaching applications. For the general teacher tier, combine online micro-courses, hands-on practical training, and school-based research, focusing on developing basic tool application and routine instructional design skills. For the trainer/researcher tier, combine theoretical study, practical guidance, and evaluation feedback, focusing on cultivating the ability to design and guide AI-specific training programs.

The training process should emphasize a "task-driven" approach, using real-world teaching problems as the impetus, requiring teachers to complete practical tasks such as AI-informed instructional design and teaching case development, thereby translating training acquisitions into tangible teaching practice outcomes. Simultaneously, leverage the interactive features of intelligent learning platforms to build communication and collaboration networks among teachers, facilitating experience sharing and collective problem-solving.

3.3 Dynamic Effect Evaluation Following Implementation

Dynamic effect evaluation must move beyond the traditional "examination + attendance" model, establishing an assessment system that combines process-oriented and outcome-oriented evaluation. Process evaluation focuses on dimensions such as teachers' training participation, task completion quality, and online interaction frequency, utilizing data automatically collected by intelligent platforms for real-time monitoring of the training process. Outcome evaluation focuses on the degree of improvement in teachers' teaching practices, assessed through classroom observations, comparative analysis of learning data, and evaluation of teaching artifacts, to gauge the impact of AI application on teaching effectiveness.

Evaluation results should be compiled into visual feedback reports, providing teachers with personalized improvement suggestions, while also serving as a basis for optimizing the training system. Based on the evaluation findings, timely adjustments should be made to training content, methods, and pacing, forming a closed-loop mechanism of "needs analysis —training implementation—effect evaluation—system optimization," ensuring the training system remains consistently relevant to teacher competency development and teaching practice needs.

4. Safeguard Mechanisms for Implementing the AI-Empowered Teacher Instructional Capability Training System

4.1 Policy Safeguards: Refining Supporting Policies for AI in Education

Policy safeguards provide the top-level support necessary for the training system's implementation. Firstly, educational administrative departments need to promulgate teacher AI literacy standards, clearly defining competency requirements for different teacher levels, thereby providing a basis for constructing the training system. These standards can reference existing teacher ICT competency frameworks for educational digital transformation to ensure alignment with national education strategies. Secondly, specific support policies for AI teacher training should be formulated, including funding guarantees, incentives for teacher participation, and policies for recognizing training outcomes. Integrating AI application capability into the assessment criteria for teacher professional title evaluation and awards can stimulate teachers' initiative to participate in training. Finally, establish a regional coordination mechanism for AI teacher training, integrating resources from universities, training/research institutions, and enterprises within the region to avoid duplicate training and resource wastage.

4.2 Resource Safeguards: Building an Intelligent Training Resource Sharing Platform

Resource guarantees must focus on integrating and sharing training resources by building a regional intelligent training resource platform. The platform should consolidate quality resources from multiple stakeholders: theoretical courses from universities, hands-on tool training from education enterprises, teaching cases from exemplary teachers, and guidance programs from research and training institutions. It must feature personalized recommendation capabilities, delivering tailored resources based on teachers' competency profiles and learning records. Additionally, it should enable achievement accumulation by incorporating outstanding teaching cases and training outcomes into the repository, facilitating resource recycling and continuous renewal.

4.3 Mechanism Safeguards: Establishing Inter-School Collaboration and Incentive Mechanisms

Mechanism guarantees must focus on collaboration among training stakeholders and teachers' sustainable development. On one hand, establish a collaborative mechanism for four types of training stakeholders, clearly defining their division of responsibilities, rights, and interest alignment. For instance, universities partner with research and training institutions to develop courses, while enterprises cooperate with schools on hands-on tool training, forming stable synergistic relationships. On the other hand, build a long-term incentive mechanism for teacher training. Beyond policy-level assessment incentives, schools should create showcase platforms for AI teaching applications, encouraging participation in teaching innovation competitions and case evaluations to enhance teachers' sense of accomplishment and fulfillment. Additionally, establish a seed teacher leadership mechanism where core teachers guide peers in AI application practices, cultivating an environment of full participation.

5. Conclusion

The construction of a training system for enhancing teachers' instructional capability empowered by AI is a process of deep integration between technological logic and educational logic. Its core lies in using systematic training design to promote the upgrade of teachers' instructional capability from "technology adaptation" to "technology integration." The training system constructed in this study, based on the principles of demand orientation, human-AI collaboration, dynamic adaptation, and literacy integration, encompasses core elements such as objectives, content, providers, and platforms. Through implementation pathways involving precise needs analysis, stratified training design, and dynamic effect evaluation, supported by policy, resource, and mechanism safeguards, it forms a closed-loop training paradigm.

Future research needs to pay closer attention to the impact of AI technology iteration on the training system and continuously optimize training content and methods. Concurrently, research on the adaptability of the training system across different regions and educational stages should be strengthened to promote its localized implementation. Only by adhering to the humanistic essence of education, centering on teacher development, and utilizing technological empowerment as a means, can the symbiotic development of artificial intelligence and teachers' instructional capability be truly realized, providing solid teacher support for high-quality educational development.

参考文献：

1. [1] Zheng Zhiyong, Song Naiqing. Research on the Construction of an Evaluation Indicator System for Intelligent Education Literacy of Primary and Secondary School Teachers[J]. China Educational Technology,2023, (12):75-83.
2. [2] Liu Dawei. The Practical Path of Generative Artificial Intelligence Assisting Teacher Teaching[J]. Teaching and Administration,2025, (20):29-33.
3. [3] Chen Dianbing, Zhu Junqi, Yang Xinxiao. Deduction of Teachers' Digital Competence from the Perspective of the 'Teacher Artificial Intelligence Competency Framework'[J]. Teaching and Administration,2025, (10):24-28.
4. [4] Huang Mei, Liu Guomin. The Development Path of Digital Literacy of Higher Vocational Teachers from the Perspective of TPACK Theory——Taking Traffic and Civil Engineering Majors as an Example[J]. Vocational Education,2024,23(12):39-44.
5. [5] Chen Dianbing, Zhu Junqi, Yang Xinxiao. Deduction of Teachers' Digital Competence from the Perspective of the 'Teacher Artificial Intelligence Competency Framework'[J]. Teaching and Administration,2025, (10):24-28.
6. [6] Gu Xiaoqing, Wang Chengliang, Wang Peijun, et al. The Mechanism, Demand, and Path of Generative Artificial Intelligence Empowering Teaching[J]. Journal of the Chinese Society of Education,2025, (04):15-22.
7. [7] Wei Fei, Zhu Zhiting. Strategies for Building Teachers' Informatization Capacity for Educational Digital Transformation[J]. Journal of the Chinese Society of Education,2022, (09):13-20.