AI and Expert Perspectives: A Comparative Analysis on Training Needs Assessment of Sudurpaschim Province

Training Need Assessment (TNA) is one of the critical functions of human resource capacity development within the organization. However, the expert-based approaches appear to be comparatively more time-consuming, costly and susceptible to subjective biases. Therefore, this study presents a comparative study between AI-based and expert-based training need assessment within a public institution context, particularly of Sudurpaschim Province, Nepal. In this study, a mixed research approach has been employed to evaluate the quality, time and cost involved in conducting both types of TNA and stakeholders' acceptance towards AI-generated TNA. The quality of both TNA reports has been assessed through Goldstein & Ford model for TNA evaluation, while ISTAM (Intelligent Systems Technology Acceptance Model) has been employed to measure stakeholder acceptability of AI-generated TNA. The findings reveal that the quality of AI-based TNA is nearly comparable with expert-based TNA in organizational analysis. However, it shows significant differences in task/KSA analysis and person analysis. Expert TNA has an overall score of 74 percent, whereas AI TNA scores 67 percent. Furthermore, AI TNA takes 50.71 percent less time and 74.34 percent lower costs, but it achieves moderate stakeholder acceptance (66.57/100). The study recommends a blended approach where AI supports efficiency while human experts provide contextual depth and empirical grounding.