Author(s): Tulika A Khare

Abstract:
Artificial Intelligence (AI) automation has become a central component of modern workplaces, influencing cognitive task demands and human performance. While automation reduces physical strain and increases efficiency, it may introduce new forms of cognitive load, such as constant monitoring, verification, and decision oversight. This study evaluates cognitive workload differences between manual task performance and AI-assisted task performance using a mixed-methods ergonomic approach. Sixty participants executed structured tasks under two conditions: manual problem-solving and AI-supported completion. Cognitive workload was measured using NASA-TLX scores, task completion time, accuracy rate, attention-shift observation, and post-task subjective feedback. Results revealed that temporal and physical demands significantly decreased under AI automation, whereas mental demand, monitoring pressure, and trust-related stress increased. The research highlights the paradox of automation—AI simplifies tasks but simultaneously elevates supervisory cognitive burden. Findings suggest the need for human-centered AI design, improved interface ergonomics, and training interventions that balance task delegation between humans and AI systems. The study contributes to cognitive ergonomics literature by providing empirical insights into the real-time cognitive effects of AI-driven work systems.

DOI: https://doi.org/10.22271/23957476.2025.v11.i3g.2033