<p>Research is being conducted on the adaptation of robots to social norms to promote their acceptance by humans. From a robot ethics perspective, the general impact of robot adaptation on the society has been qualitatively discussed. Today, highly socialized robots are not used widely; hence, their impact has not been examined quantitatively. Therefore, this study investigates robots’ adaptation to group norms based on an opinion dynamics methodology and simulates its impact on human–robot societies by agent-based simulation method. Using different parameters related to the strength of the influence of robots on human opinions and thresholds at which people tolerate the opinions of others, this study quantitatively analyzes the impact of robots’ adaptation to group norms on consensus building in a society. Results show that robots promoted convergence of diverse opinions in societies over time even when robot’s social influence on human’s opinion was less than human’s social influence on another human. Additionally, in societies where humans are less likely to tolerate differences in opinion, the majority opinion in human–robot societies will be different from the one that will be originally formed in human societies. Based on qualitative suggestions from the robot ethics field, simulation results can be discussed based on quantitative data regarding the impact of robots and artificial intelligence on the society. In addition, this study may help build a foundation for the robot design considering the human characteristic of conforming to others.</p>