Automated decoding of facial expressions reveals marked differences in children when telling antisocial versus prosocial lies.

•Children aged 6–11 told either an antisocial lie or a prosocial lie.•We used computer vision technology to automatically code facial expressions.•Children’s facial expressions differed in intensity depending on the lie told.•The intensity of blinking, lip puckering, and lip raise differed per lie-type.•Children showed more joy while telling a prosocial lie than an antisocial lie.