Junwei Lin, Zihan Zhang, Zinuo Meng, & Huan Zhang
School of Government
Beijing Normal University, Beijing, China
Correspondence: Huan Zhang, zhanghuan@bnu.edu.cn

---

While Artificial Intelligence (AI) is frequently championed as a transformative tool for public services, its frontline implementation is often characterized by “agitation” rather than seamless integration. Moving beyond the cognitive appraisal of Technology Acceptance Model, this study investigates the sustained state of uncertainty and negotiation among frontline public service deliverers. Drawing on 19 in-depth interviews with frontline public service deliverers in China, this study develops a Context-Value-Behavior framework to deconstruct this complexity. Our findings reveal a fundamental tension: the algorithmic logic of AI, which prioritizes standardized efficiency, directly clashes with the professional logic of frontline public service deliverers, which is rooted in contextual intelligence and human-centric directions. They employ a range of behavioral strategies—from deep integration to deliberate disengagement— while constantly wrestling with the trade-offs between systemic efficiency and core professional values under constrained resources. The primary contribution of this study is to extend the Technology Acceptance Model by situating it within the messy realities of frontline service delivery, demonstrating that technology adoption is an ongoing, agentic process of co-evolution. We conclude that for AI to be truly enabling, policy must recognize, actively engage with, and adapt to the lived experience of frontline agitation.

Keywords: Frontline public service deliverers, Artificial Intelligence, Agitation, Context-Value-Behavior Framework, Technology Acceptance Model