Key facts
- KPMG rebuilt its early careers training program for nearly 1,000 interns.
- AI now handles routine analysis and cross-checking that previously occupied junior professionals.
- KPMG's new training emphasizes scenario work and direct coaching.
- Interns reported direct coaching as the most valuable learning method.
- Concerns about AI limiting critical thinking were more prevalent than job displacement fears among interns.
KPMG has undertaken a significant overhaul of its early careers training, adapting to a landscape where artificial intelligence now performs the routine, low-stakes tasks that historically served as the primary method for developing junior professionals' judgment. The firm's new approach, observed by The Wall Street Journal and Fortune, centers on scenario-based exercises and direct human coaching to impart skills that were once absorbed organically through repetitive work.
For decades, graduates entering fields like auditing spent their initial years performing high volumes of repetitive tasks, such as testing samples and checking invoices. This process, while tedious, was instrumental in building competence and judgment. However, KPMG's internal assessment indicates that AI now handles approximately three-quarters of this 'middle to middle' work, necessitating a new curriculum.
KPMG's own July survey of 906 US summer interns revealed that 57% found direct coaching from a person to be the most valuable learning method, significantly more than any other option. Conversely, AI-assisted learning ranked last at 3%. A notable concern among interns was the potential for over-reliance on AI to limit their critical thinking, with 43% citing this as their top worry, far outweighing the 5% concerned about AI taking their jobs.
Interns like K-Linh Nguyen highlighted the difficulty in teaching the nuanced skill of discerning when AI outputs are incorrect, suggesting that such judgment requires lived experience. KPMG acknowledges the uncertainty surrounding whether judgment can truly be taught in a classroom setting, particularly when compared to the experience of making mistakes on real client work and receiving correction from senior professionals. Field studies by KPMG with the University of Texas at Austin suggest that guiding and checking AI is more critical than existing knowledge for early-career professionals.