Organizational Interfaces and Autonomous Agents

Abstract

The history of organizational research is abundant with organizational interfaces: human-to-human, machine-to-machine, and human-to-machine. Existing research on coordination treats humans and machines as distinct. However, in recent years, the emergence of autonomous agents is rendering the distinction between human and machine less salient. As the line between human and machine agency blurs, we propose a framework of agent-to-agent interfaces. The framework involves four general types of agent-to-agent interfaces: standard, adaptive, generative, and autonomous. Each type of interface requires a different perspective for organizational action, and we propose design principles, reference domains, and modes of adaptation for each.