Human- centred AI in elderly care: alleviating loneliness, promoting social engagement, and easing cognitive burden

The global aging population is confronted with an escalating issue of social loneliness, which exerts a detrimental impact on the physical and mental health of the elderly. This research delves into the utilisation of AI virtual agents as a means to mitigate loneliness, augment social participation, and decrease cognitive load among senior citizens. A control experiment involving 136 participants aged 60 and over was carried out. In this experiment, the experimental group interacted with AI agents, while the control group engaged in traditional social interactions. By employing the UCLA Loneliness Scale, Social Participation Scale, and NASA – TLX Cognitive Load Scale, the results revealed remarkable reductions in loneliness (p < 0.01), a significant increase in social participation (p < 0.01), and a notable decrease in cognitive load (p < 0.01) within the experimental group. Moreover, a robust negative correlation was identified between loneliness and social participation (r = –0.938, p < 0.01). These findings indicate that AI virtual agents can serve as an effective tool to alleviate the loneliness of the elderly and enhance their overall well-being.