Transforming Talent for Agentic AI-Enabled Services: Developing and Critically Testing a Strategic Human Resource Reskilling Framework

Abstract

Agentic artificial intelligence (AI) is changing service work by enabling systems to plan, coordinate tasks, use tools, and perform actions with limited human supervision. This development changes how employees and AI systems share judgment, responsibility, and control. However, limited empirical research explains how organizations can prepare employees for effective work with Agentic AI. This study develops and evaluates an Agentic AI Talent Transformation Chain by integrating dynamic capabilities theory and self-determination theory. The model links perceived organizational dynamic capabilities, the AI reskilling environment, AI-learning basic psychological need satisfaction, AI reskilling adaptation and mastery, and perceived Agentic AI-enabled service performance. Data collected from professionals working in the Indian banking sector. A total of 161 responses were retained for analysis. A disjoint two-stage PLS-SEM approach was used to assess the measurement and structural models. The pooled analysis showed significant direct and indirect relationships. However, further analysis identified weak reliability in the two-item dimensions, poor discriminant validity, weak redundancy validity for service performance, and substantial response-block heterogeneity. After controlling for the response-block effect, most structural relationships became non-significant, and the full serial pathway was no longer supported. The findings therefore provide preliminary rather than confirmatory evidence for the proposed framework. The study contributes a theoretically grounded model for Agentic AI reskilling and highlights the importance of measurement quality and verified data provenance in emerging AI workforce research.