AI Agents Framed as Coworkers Impair Human Performance

Framing artificial intelligence agents as "coworkers" or "employees" rather than software tools can negatively impact human performance and lead to unrealistic expectations, according to research by Boston University business professor Emma Wiles. Wiles's study found that managers caught 18% fewer errors when the work was attributed to an agentic "AI employee" compared to a chatbot. This suggests that the nomenclature used for AI tools significantly influences human interaction and effectiveness.

This approach to AI integration is gaining traction in Silicon Valley, with companies like Nvidia, Microsoft, OpenAI, Anthropic, and Google developing tools that manage teams of AI agents. These agents are often advertised as digital colleagues possessing human-like flexibility and cognitive abilities. Wiles's study revealed that nearly a third of 1,261 surveyed managers reported their companies already describe AI agents as employees, with 23% even including them on organizational charts.

While the technical capabilities of agentic AI, which are AI tools designed to operate in loops to achieve specific goals, have demonstrably improved for complex tasks, the leap to portraying them as coworkers or employees is problematic. This framing can create a disconnect between the AI's actual capabilities and the expectations placed upon it, potentially disadvantaging human employees who are tasked with overseeing these systems. Wiles's research further indicates that this framing can invert the perceived hierarchy, leading individuals to feel less in control when interacting with AI presented as an employee.

The implications of this framing extend to how human employees perceive their roles and responsibilities when working alongside AI. By anthropomorphizing AI agents, companies risk blurring the lines between human and machine contributions, potentially leading to a devaluation of human oversight and critical error detection. The study highlights the importance of clear and accurate communication regarding the nature and limitations of AI tools in the workplace to foster effective collaboration and maintain human accountability.