Modern A.I. systems excel at many tasks, but lack essential capabilities for interacting and cooperating with other intelligent being. We investigate how A.I. technology can be advanced to a level of social and cooperative intelligence required for successful human-agent collaboration or teamwork. Our research bridges A.I., Cognitive Science and Human-Agent/Robot Interaction: We study human cooperation and communication and develop A.I.-based models for corresponding behavioral and cognitive abilities, e.g., Theory of Mind, collaborative decision-making, nonverbal communication, or multimodal dialog. We apply those models in novel human-agent interaction scenarios involving socially assistive robots, conversational agents or collaborative A.I. agents .