The Enterprising and Elusive Prospects of Human-AI Collaboration

Abstract

The exponential advances in generative AI and agentic technologies have presented organizations with an unprecedented opportunity to seek and attain competitive advantages. Some aim to do so through product and service innovation, while others have begun to pursue operational efficiency and productivity. There is a pivotal role that human-AI collaboration can play in enterprise artificial intelligence design, development, and deployment today. In this chapter, we synthesize developments in human-AI collaboration over the last decade and explore how organizations can design AI systems that augment rather than replace human capabilities to achieve optimal experiential and performance-related outcomes for different stakeholders. Drawing from empirical work across different domains, we analyze the key factors that shape trust and reliance and determine successful human-AI collaboration. This includes various human factors, task factors, and AI system factors and the complex interplay between them in different configurations. This synthesis reflects the importance of broadening the spectrum of metrics for evaluating human-AI collaboration (e.g., by considering stakeholder values). We discuss promising ideas to address common challenges such as under-reliance or over-reliance on AI systems. We argue for broadening the lens of human-AI collaboration to consider the AI supply chain and the underlying value chains to ensure the responsible design of enterprise AI. Our insights collectively suggest that enterprise AI will benefit from the human-centered approach while creating collaborative workflows and human-AI configurations that can augment and complement humans with the computational power and scalability of AI.