This essay explains complementary frameworks for understanding and managing AI in usage contexts. In contrast with broad generalizations about the nature and impact of AI, those frameworks focus on specific AIbased digital agents used by people and/or machines performing purposeful activities in business, home, or societal environments. The agent responsibility (AR) framework helps in describing roles and responsibilities of specific AI-based digital agents in their usage contexts. The agent evaluation (AE) framework identifies six criteria that different stakeholders might use for evaluating AI-based digital agents.