Kids Trust Unreliable Humans More Than Robots — and Age Explains Why

One key skill children need today is judging which information sources they can trust and developing critical thinking.

Researchers from the National University of Singapore found that children were more likely to trust an unreliable human informant than an unreliable robot. But that was only one of their observations.

The research was led by Dr. Li Xiaoqian and her academic supervisor, Professor You Wei Quin. Together with their team, they found that “children don’t just trust whoever names things for them—they trust people who have proven reliable in the past.” They say this selectiveness in social learning reflects how children come to understand what makes a source trustworthy.

“The question is how young children decide when to learn and whom to trust,” Dr. Li Xiaoqian said.

How the research was conducted

The team tested preschoolers from ChildFirst, Red SchoolHouse, and Safari House in Singapore. The average age of participants was 4.58 years, which the researchers used to divide the group into younger and older children.

Participants interacted alternately with either a human or a robot informant (a humanoid robot with a robotic-sounding voice). These informants gave either accurate or inaccurate definitions of various objects.

To measure trust, the researchers watched how the children responded to new information from those informants, according to Earth.

What the team discovered

The results showed that children trusted informants—whether human or robotic—who had previously provided accurate information. They were reluctant to trust informants who had given incorrect information in the past, especially when the informant was a robot.

Younger children were more likely to trust an unreliable human than an unreliable robot. Older children, in contrast, tended to distrust unreliable informants whether the source was a robot or a human.

“These findings show that younger and older children use different selective-trust strategies, especially in how they weigh an informant’s past accuracy versus identity cues when deciding whom to trust,” Dr. Li said.

She said that as children get older, they increasingly rely on cues about reliability to guide whom they trust.

Previous studies have shown that children evaluate an informant’s reliability using factors such as age, familiarity, and speech. Younger children tend to focus more on identity cues, while older children prioritize the content of the information.

This study is significant because it looked at how children perceive information from both humans and robots, and how trust-based behavior develops in early childhood.

Professor You Wei Quin said understanding this gives insight into how trust and social learning develop in children growing up surrounded by diverse sources of information, including artificial intelligence.

The results of the study were published in the journal Child Development.