The human gaze speaks to children, who remain insensitive to that of humanoid robots

Very young children (even as young as 3 years old) can read the eyes of a human gaze, grasping preferences and
intentions, whereas they do not recognise this type of non-verbal communication in the gaze of a humanoid.
This is what was discovered in a study published in the International Journal of Child Interaction, coordinated by Antonella Marchetti, director of the Department of Psychology at the Università Cattolica, Milan campus, and of CERITOM (Centre for Research on Theory of Mind and Social Skills in the Life Cycle) in collaboration with scholars from Tokyo and Osaka, and with colleagues Davide Massaro, Cinzia Di Dio, and Federico Manzi of the Università Cattolica, Milan campus.

The research involved Italian children aged 3-5 years to explore how a human gaze or a robot gaze can elicit different impressions in the minds of young children. The test consisted in having the children see the gaze of a person and a humanoid robot on an object, assessing whether they are able to understand which object is 'preferred' by the agent looking at it.
The results show that the children interpret the human gaze as a meaningful signal: if an individual looks at an object, the children tend to think they like that object. The same does not happen, however, when a humanoid robot looks at the object. In that case, the gaze is not sufficient for children to attribute a true preference to the robot.

In short, children use the human gaze to 'read' desires and intentions, whereas they struggle to do the same with the robot. Moreover, the gaze - human or robot - does not seem to change the children's personal preferences: it helps them understand what the other person likes, but does not necessarily change their own preferences.

'This does not mean that robots cannot play an educational or social role,' explains Professor Marchetti, 'but it does suggest that it is not enough to imitate a single human signal, such as a gaze, in a robotic artefact to make it truly communicative in the eyes of a child. Designing intelligent robots and technologies for children requires richer, more natural and developmentally appropriate interactions: made of words, gestures, reciprocity, context and shared presence. This is reinforced by the fact that even the human gaze alone is not sufficient to exert clear transformative effects on child preferences'. "These data are particularly topical in the debate on artificial intelligence," she continues. "Many AIs today speak, respond and suggest, but our results highlight that, especially for children, communication is not only through words: presence and shared context also count. In this perspective, an AI integrated in physical supports, the so-called embodied AI that finds in humanoid social robotics one of its most complete expressions, represents a crucial dimension to understand how children attribute mental states (such as intentions, beliefs, preferences) also to technologies'.

These results also have significant application implications, particularly in the field of autism, where gaze and shared attention represent crucial psychological dimensions of social-communicative development and can be particularly fragile. In this very field, humanoid robots are increasingly being studied as support tools for rehabilitation interventions centred on these skills. Understanding under what conditions a child interprets a robot's gaze as an intentional signal can therefore help design more targeted, natural and developmentally sensitive interventions.
The ROBIN project (ROBot-based neuropsychomotor INtervention to promote imitation skills in young children with autism spectrum disorder), financed by the Ministry of Health as part of the Finalised Research that will begin in June 2026, is also part of this research direction, anticipates the professor. The project is led by the Don Carlo Gnocchi Foundation and the CeRiToM of the Catholic University of the Sacred Heart, involved as a research group on the role of the gaze and psychological processes in these forms of intervention. The project envisages interventions with a humanoid robot to promote imitation skills, which also involve understanding the robot's gaze and its communicative meaning.