<![CDATA[Artificial intelligence capability is increasingly viewed as an important organizational resource in public-sector digital transformation, yet the possession of AI-related resources does not necessarily translate into sustained employee-level use and performance. This article develops an accountability-bounded enactment framework explaining how organizational AI capability becomes consequential through employee adaptation and work practices. Using a theory-synthesis approach supported by structured thematic and targeted literature searches, the study integrates research on AI capability, organizational learning, employee unlearning, technology enactment, AI use, sustainable employee performance, and public-sector accountability. The framework generates three main theoretical insights. First, employee learning and unlearning are not treated merely as parallel adaptive mechanisms: employee unlearning is also proposed to condition whether newly developed AI competence becomes task-integrated use. Second, where sanctioned learning opportunities remain available, procedural rigidity is expected to constrain capability-driven unlearning more strongly than capability-driven learning. Third, accountability clarity strengthens the positive relationship between task-integrated AI use and employee sustainable performance, while not eliminating other potential costs of AI use such as cognitive burden, dependence, or technostress. These relationships are theoretically derived rather than empirically tested. The framework contributes a cross-level explanation of AI capability enactment in public-sector organizations by showing how competence development, routine revision, task-integrated AI use, and accountability conditions jointly shape whether organizational AI capability produces sustainable employee outcomes.]]>
Copyrights © 2026
