Related Experiment Videos
The cognitive cost of artificial intelligence on ageing workforce: Analyzing age-related effects on mental workload
Nurul I Abd Rahman1,2, Muhammad Ni Ismail1, Siti A Ahmad1
1Department of Mechanical and Manufacturing Engineering, Faculty of Engineering, Advanced Manufacturing Research Centre (AMREC), Universiti Putra Malaysia, Selangor, Malaysia.
Abstract:
BackgroundThe current world demographic reality necessitates a critical re-evaluation of how technological tools, specifically those powered by artificial intelligence (AI), interact with the varying cognitive profiles of a ageing workforce.ObjectivesThe objective of this study is to evaluate the spatial distribution of mental workload and objective task performance across AI-assisted and non-AI-assisted conditions in a computer-based environment.MethodAn experimental study involving 30 older workers compared standardized office tasks under non-AI-assisted (Segment A) and AI-assisted (Segment B) conditions. Mental workload was assessed using NASA-TLX, Heart Rate Variability (HRV), and EEG-based Beta-Alpha Ratios (BAR) across frontal (F3, F4), parietal (P3, P4), and occipital (O1, O2) regions. Performance in document preparation (Task 1) and arithmetic operations (Task 2) was evaluated using percentage-based accuracy, efficiency, and overall performance scores.ResultAI intervention improved overall task performance but was accompanied by an apparent performance-workload paradox, whereby enhanced output coincided with increased neurophysiological and systemic strain. Descriptively, NASA-TLX Overall Perceived Workload decreased from 54.05 to 44.91 during complex tasks, while HRV declined by 48.68% in Task 1, indicating acute autonomic strain. EEG revealed increased BAR in the frontal (F3: 29.2%), parietal (P3), and occipital (O1) regions, suggesting heightened attentional and sensory demands associated with monitoring and verifying AI-generated suggestions.ConclusionAI redistributes rather than eliminates workload, increasing executive verification and visual monitoring demands. Future AI systems for older adults should prioritize cognitive ergonomics to minimize executive burden and sensory strain, preventing premature cognitive fatigue while enhancing task performance and supporting sustainable human-AI interaction.