KPMG and researchers from the University of Texas at Austin's McCombs School of Business published findings from a field study of 523 early-career professionals who completed business-specific tasks using a domain-built AI agent. The study sorted participants into three behavioral groups — AI apprentices, AI delegators, and AI amplifiers — based on how they worked with the agent on higher-stakes professional tasks requiring judgment, domain expertise, and decision-ready outputs. The authors report that foundational attributes such as critical thinking ability, AI literacy, and domain knowledge were not clear predictors of which group a participant fell into. The write-up, authored by Ashish Agarwal, Anitesh Barua, Anu Puvvada, Fangchen Song, and Wen Wen, appeared in Harvard Business Review on July 23, 2026, and frames implications for how entry-level work is designed and how junior roles are hired for.