Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Humans behave pro-socially and the degree to their pro-social behavior depends crucially on social norms. I ask whether artificial intelligence chatbots exhibit pro-social behavior which is influenced by information on human norms. I show that it is. AI is asked to make an allocation decision in a Dictator Game. Treatments are included with either descriptive and injunctive norms. When the social norm is to be more generous, AI’s willingness to give to others improves. AI is equally responsive to descriptive and injunctive norms.