The clean-energy transition is simultaneously a technological, environmental, and workforce transformation. Although research increasingly links artificial intelligence (AI) with green human resource management (GHRM), existing work largely emphasizes efficiency, green behavior, or sustainable performance, leaving the workforce consequences of decarbonization insufficiently theorized. This conceptual article develops a capability-conversion framework explaining how AI-driven GHRM can support a just energy transition through workforce reskilling. Integrating ability-motivation-opportunity theory, the natural resource-based view, and dynamic capabilities theory, the framework positions AI-driven GHRM as an organizational capability that converts transition intelligence into green workforce capabilities through reskilling, redeployment, participation, and responsible performance management. It distinguishes sustainable performance from just-transition outcomes and argues that environmental gains are incomplete when they are achieved through exclusion, occupational downgrading, inequitable access to training, or opaque algorithmic decisions. Responsible AI governance-comprising human oversight, transparency, contestability, employee voice, and distributive justice-is theorized as a cross-cutting boundary condition. Seven propositions specify the relationships among transition pressures, AI-driven GHRM, reskilling, green workforce capability, sustainable performance, and just-transition outcomes. The article contributes to GHRM and AI-HRM scholarship by shifting attention from technology adoption to capability conversion and from aggregate sustainability to the distribution of workforce opportunities, risks, and benefits. A future research agenda outlines longitudinal, multilevel, comparative, and algorithm-audit designs for empirical testing.