As AI agents acquire capacities to decide autonomously, we switch from being omnipotent users of intelligent machines (e.g., Google Translate) to having to make decisions with or beside them in social interactive settings (e.g., sharing the road with self-driving cars). The impact of this on people’s choice behaviour and, in turn, on the desirability of outcomes of human-AI interaction is yet unknown. In this project we investigate whether and when humans will cooperate with AI for the attainment of mutually beneficial and efficient results.