GenAI Empowering Vocational Education Ecosystems: A Digital Innovation Ecosystem Theory Perspective

Abstract

The rapid development of generative artificial intelligence (GenAI) is propelling vocational education from instrumental digitalization toward ecological reconstruction. Existing research has largely focused on GenAI's value in teaching assistance, resource generation, and learning support, but has insufficiently addressed how it reshapes multi-actor relationships, organizational structures, and governance logic within vocational education. This paper introduces the Digital Innovation Ecosystem (DIE) theory and analyzes the intrinsic mechanisms through which GenAI empowers the ecological reconstruction of vocational education across five dimensions: innovation actors, innovation structure, innovation institutions, innovation functions, and innovation evolution. The study argues that GenAI, as an intelligent agentic element with connecting, translating, and coordinating functions, drives teachers, students, enterprises, industry organizations, and platform institutions to form a new type of human-machine co-creation collaborative relationship. By interconnecting data infrastructure, linking intelligent platforms, and coordinating task networks, GenAI propels vocational education's structure from resource aggregation to process coupling. Relying on data governance, algorithmic ethics, and human-machine coordination institutions, it ensures safe and orderly technology application, and releases educational value through practical knowledge translation, higher-order competency generation, and industry-education demand matching. At the same time, GenAI's empowerment of vocational education still faces real impediments: insufficient actor collaboration, loose structural connections, lagging institutional supply, constrained functional release, and weak evolutionary mechanisms. Accordingly, this paper proposes five practical pathways: layered activation of actor participation to resolve the coordination deficit; a 'small but precise' strategy to advance data infrastructure construction; establishing a 'graded and scenario-based' institutional governance framework; reconstructing an embodied experience-oriented task resource system through 'human-machine collaborative practical training'; and building a progressive evolutionary mechanism of 'small-step piloting - process reengineering - reciprocal co-creation'.