Embodied AI as Companion: How Loneliness, Gender, and Culture Shape Attitudes towards AI and Robots

Loneliness among university students is a widespread problem, affecting mental health and academic performance. While AI-driven chatbots offer digital companionship, their lack of embodiment may limit their ability to foster meaningful connections. In contrast, social robots, a form of embodied artificial intelligence (EAI), offer physical presence and non-verbal interaction. This study examines how social (lack of a broad social network) and emotional loneliness (lack of close emotional bonds) influence students' attitudes towards AI and robots, considering gender and cultural background. A cross-sectional online survey (N = 250) was analysed using Partial Least Squares Structural Equation Modelling (PLS-SEM). Results show that social loneliness reduces negativity towards robots, but does not increase preference for their companionship, while emotional loneliness increases scepticism towards AI, but does not significantly affect attitudes towards robots. Gender moderates these effects, with emotionally lonely women expressing greater negativity towards robots and socially lonely women expressing greater openness. Cultural background does not moderate these relationships, but individualistic participants show less positivity towards AI and robots. Frequent interaction with AI and robots correlates with more positive attitudes, suggesting that familiarity promotes acceptance. These findings provide insights into how loneliness and demographics shape attitudes towards AI and inform the development of EAI interventions for student well-being.