Why You Shouldn't Treat AI Agents Like Employees (hbr.org)

As AI adoption accelerates, some organizations are framing AI systems as "employees" to make the technology more relatable. However, a recent study reveals significant pitfalls in this approach. Treating AI as a colleague can blur accountability and decrease the quality of work. In an experiment with over 1,200 managers, it was found that framing AI as an employee led to a 9% decrease in personal accountability and an 18% drop in error detection during document reviews, as participants felt less engaged in scrutinizing AI-generated outputs. This anthropomorphization may create an illusion of shared responsibility, enabling individuals to shift blame to AI for errors, which undermines professional identity and trust in the technology. This shift from AI as a tool to an employee-like status poses critical challenges for organizations. While humanizing AI aims to ease integration, the findings indicate it does not effectively enhance intent to adopt AI technologies. Rather than increasing confidence in using AI, it potentially erodes trust and creates uncertainty regarding job roles among human employees. Leaders must reconsider how to integrate AI into workflows, ensuring clarity regarding accountability, and redefining employee roles to foster effective human-AI collaboration while maximizing the value AI can provide.