Stakeholder Engagement in AI-Supported Education: A Case Study of a Primary School in China

• In Event: Building Peace in Primary Schools: Life Skills, Stakeholder Engagement, and AI-Supported Education Across Global Contexts

Proposal

Introduction

The growing integration of artificial intelligence (AI) into primary education has become a global trend, raising questions not only about pedagogy but also about governance and cross-sector collaboration. As countries and international organizations release frameworks for AI in education, the issue has clear comparative and international relevance.

Nevertheless, much of the existing scholarship has concentrated either on higher education (Zawacki-Richteror et al., 2019; Crompton & Burke, 2023) or on classroom-level pedagogical applications of AI in primary schools (Pardamean et al., 2022; Yim & Su, 2025). What remains underexplored is how primary schools mobilize and coordinate external resources—policy, technologies, academic expertise, and parental engagement—to implement AI sustainably. Importantly, the implementation of AI in schools is not simply a matter of technology adoption but rather a socio-technical process involving multiple actors, diverse resources, and complex mechanisms of collaboration. Thus, effective adoption depends on the school’s ability to negotiate and orchestrate collaboration across sectors. This highlights why examining external resources and stakeholder relationships is central to understanding the trajectory of AI in education.

China provides an especially context. National strategies, such as Guide to General Education in Artificial Intelligence for Primary Schools (2025) explicitly encourage experimentation while mandating attention to ethics and collaboration. At the municipal level, Shanghai’s Action Plan for Advancing AI-Empowered High-Quality Development of Basic Education (2024–2026) emphasizes building a safe, collaborative, and efficient AI education ecosystem. These policies emphasize that the application of AI necessitates collaboration among various stakeholders. Only by establishing an effective multi-agent cooperation mechanism can resources from all parties be fully integrated to foster the development of AI-supported education.

Against this backdrop, this study addresses three guiding questions: First, what external resources are mobilized by primary schools to develop AI-supported education? Second, how do different stakeholders interact, collaborate, and negotiate throughout the process of AI integration? Lastly, what mechanisms facilitate value co-creation, and what tensions or conflicts arise among stakeholders?

Theory

This research is guided by stakeholder theory, which posits that organizations operate within networks of actors whose resources, interests, and power influence outcomes (Freeman, 2010; Parmar et al., 2010; Matthews et al., 2025; Yihong et al., 2024). In the Chinese context, the development of AI in education cannot be sustained by schools alone but depends on dynamic interactions among government, technology enterprises, universities, as well as parents. Therefore, stakeholder theory provides a useful analytical framework for unpacking how these actors contribute to, benefit from, and negotiate around the use of AI in primary education.

Methods

A qualitative case study approach was employed, suitable for exploring complex socio-technical processes in depth. The research was conducted in a private primary school in urban Shanghai, China, designated as a demonstration school for curriculum development. Through collaboration among stakeholders, AI has been integrated into curriculum innovation and interdisciplinary teaching, significantly improving instructional quality and efficiency. The school thus represents a distinctive and exemplary case of how external resources can be mobilized to advance AI-supported education.

Data collection employed a multi-method strategy: semi-structured interviews with six participants from key stakeholder groups (teachers, administrators, enterprise representatives, and university faculty), document analysis of school plans, policy documents, cooperation agreements, and curriculum materials, as well as classroom observations to examine the integration of AI tools into practice. Data analysis followed thematic coding guided by stakeholder theory. Triangulation across interviews, documents, and observations enhanced reliability, while discrepant perspectives revealed tensions within the stakeholder network.

Findings

The findings reveal a nuanced picture of multi-stakeholder engagement in AI-supported education. First, the case school has developed distinctive partnerships with technology enterprises, universities and parents: enterprises deliver tailored AI tools and technical support; universities provide teacher training and co-develop curricula; and parents contribute critical feedback. This reflects the principle that effective AI adoption requires broad cross-sector collaboration (Holmes et al., 2023).

Second, external resources are mobilized through four pathways: (a) policy, (b) technology and training, (c) academic collaboration, and (d) parental engagement. These pathways illustrate how schools translate stakeholders’ resources into practical support, reinforcing resource integration as a key governance function (Freeman, 2010).

Third, collaboration among stakeholders operates through mechanisms of resource complementarity, co-governance, and value co-creation. For instance, enterprises gain pilot environments for testing products while schools benefit from technological support; teachers and university researchers co-develop AI curricula; and parents’ concerns about screen time and privacy lead to application adjustments. This resonates with Luckin’s (2016) argument that sustainable AI integration emerges from iterative partnerships rather than unilateral implementation.

Finally, tensions and risks are evident: teachers face increased workloads and insufficient training; parents remain concerned about privacy; and enterprises, while providing vital resources, sometimes prioritize commercial interests, creating conflicts with educational values. Moreover, power imbalances risk marginalizing teachers’ voices in governance, highlighting stakeholder theory’s insight into the asymmetry of influence. These risks underscore Holmes et al.’s (2022) call to view AI education as a socio-technical ecosystem where conflicting priorities must be carefully balanced.

Conclusion

The study concludes that the development of AI in primary education follows a dynamic trajectory of “multi-stakeholder participation – resource integration – value co-creation.” This indicates that sustainable AI adoption in schools requires not only technological readiness but also effective governance, and participatory mechanisms. By situating AI-supported schooling within a stakeholder framework, the research extends stakeholder theory into the underexplored context of primary education, demonstrating its utility in analyzing resource mobilization, power dynamics, and collaborative governance within socio-technical systems.

Moreover, the study provides evidence that governance challenges identified in international debates—such as balancing commercial interests with educational values, sustaining collaboration, and protecting privacy, (Pedro et al., 2019; Eden et al., 2024) —are not abstract but lived realities in schools. Practically, this implies that schools should build internal capacity while strategically managing partnerships; policymakers should ensure equitable and sustainable support; enterprises must align innovation with educational values; and universities should reinforce both AI’s pedagogical role and ethical application. In this sense, the findings extend beyond a single case: they illustrate how global issues on ethics, collaboration, and governance are negotiated in primary school, thereby contributing to international comparative scholarship on the socio-technical conditions under which AI can enhance, rather than undermine, human-centered education.

Authors

• Jianqiao Lai, East China Normal University

• Li Deng, East China Normal University

• Yan Gao