AI-Driven Training, Employee Skill Development and Employee Outcomes: Evidence from Public-Sector Energy Sector

Abstract

It’s not very clear how AI-based training (AIDT) improves employee outcomes and under what organizational conditions. The current study looked into the relationship between AIDT and decision-making capability (DMC), performance enhancement (PE), and work productivity (WP) among employees of public sector electricity distribution companies (DISCOs) in Pakistan, including employee skill development (ESD) as a mediator and organizational culture (OC) as a moderator. Partial least squares structural equation modelling (PLS-SEM) was used to analyze survey data from 423 respondents, with three parallel models based only on different outcome variables. The measurement quality was good (HTMT ≤ 0.44; item loadings 0.74-0.83). Bootstrapped results show a positive relationship between AIDT and ESD (β = 0.22, p < 0.001), and ESD in turn is related to DMC (β = 0.27), WP (β = 0.26), and PE (β = 0.20). AIDT also has a direct effect on WP (β = 0.22) and PE (β = 0.10); there is no direct effect on DMC (β = 0.09, p = 0.082), which suggests that skills fully mediate the effect on DMC and partially mediate it for WP and PE. ESD is significantly related to each outcome from OC (β = 0.28; 0.18-0.33, respectively), while the interaction of AIDT × OC is not significant in any model. The models explain 14% of the variance in ESD and 17-22% in the outcomes. These results indicate that AI-driven training has positive direct effects on outcomes and promotes the development of skills, and a supportive culture positively affects the outcomes, but it doesn’t change the strength of the training effect. The results provide a basis for HR managers and policymakers in the power sector to plan AI-powered training.