AI-Driven Performance Appraisal in India: A Socio-Technical Analysis of Efficiency, Bias, and Employee Trust

Abstract

India has emerged as one of the fastest adopters of AI-driven performance appraisal systems; nonetheless, little empirical data exists on how they can be applied within hierarchical-based, linguistically diverse, and culturally embedded organisational settings. The gap is especially problematic because algorithmic decision-making is becoming more and more influential on employee assessment and career advancement.

The proposed study is based on a convergent mixed-method design to address the issue of adoption, effectiveness, and ethical implications of AI-based appraisal systems among Indian organisations. Structural Equation Modelling (SEM) was used to analyse quantitative data collected from 312 HR professionals and employees in five fields, which include IT/ITES, manufacturing, banking, e-commerce, and pharmaceuticals, and thematic analysis was applied to the qualitative data collected on six organisational case studies.

The results show that AI-based appraisal systems can lead to better operational performance, with the time spent on the evaluation cycle being decreased by 43 percent and the consistency of cross-location ratings also being elevated. Moreover, the regulatory system in India is still more underdeveloped compared to international ones, especially in matters of transparency, explainability, and accountability of algorithmic decision-making. This paper presents one of the first large-scale empirical analyses of algorithmic bias in AI-enabled HR systems in an emerging economy, which offers both theoretical progress and practical information on how an AI-based performance management practice can be designed to be transparent, fair, and context-sensitive.