Reskilling and Leadership for AI-Enabled Public Administration: An Integrative Review and a Workforce Capability Framework

Governments are adopting artificial intelligence (AI) to deliver services, process cases and inform policy, but the success of these initiatives depends less on the technology than on the capabilities of the public workforce that must use, oversee and explain it. Evidence on how AI reshapes tasks suggests that routine cognitive work is most exposed, while demand grows for analytical, interpersonal and ethical judgement. Yet research on public-sector AI has concentrated on technology adoption and governance, and far less on the reskilling and leadership needed to make adoption work. This paper addresses that gap through an integrative review of literature from public administration, labour economics, leadership and learning science, supplemented by policy sources. The review identifies three capability clusters that AI-enabled public administration requires: digital-technical capability, human-relational capability and adaptive leadership capability. It also finds that evidence-based learning designs, particularly spaced, retrieval-based microlearning embedded in work, offer a scalable route to building these capabilities across large civil services. Drawing on dynamic capabilities and transformational leadership theory, the paper proposes the AI-Ready Public Workforce Capability Framework, which links institutional enablers, capability clusters and learning architecture to public value outcomes. Four propositions and practical recommendations are offered for public managers and training institutions, with reference to India’s Mission Karmayogi.