Junior Employees Show Uneven AI Performance in High-Stakes Work

Entry-level knowledge workers face a fundamental shift in how they build careers. Tasks that once served as training grounds—analytical assignments, information synthesis, research—are increasingly handled by AI systems. Yet new research suggests not all junior employees navigate this transition equally well.

A collaboration between KPMG and researchers at the University of Texas at Austin's McCombs School of Business examined how individual contributors create value beyond what AI can deliver in real organizational settings. The study focused specifically on higher-stakes professional work that requires judgment, domain expertise, and decision-ready outputs—not simple task completion.

The rising AI baseline

The research team notes that AI capabilities continue to advance rapidly, creating what they call a "resetting baseline." As models improve, the standard for acceptable output rises correspondingly. What qualified as strong work six months ago may now represent merely adequate performance when AI can produce similar results.

This dynamic places particular pressure on junior employees, whose roles historically centered on tasks now vulnerable to automation. The question is no longer whether AI can handle analytical work, but how individual workers differentiate themselves when competing against increasingly capable systems.

Why it matters

Organizations investing in AI adoption need to understand which employees can effectively complement these tools versus those who struggle to add value beyond machine output. The variation in performance has direct implications for hiring, training, and workforce planning. Companies that assume uniform AI proficiency among junior staff risk misallocating resources and missing opportunities to develop high-potential talent who excel at human-AI collaboration.

The findings also signal a shift in what constitutes valuable entry-level skills. Traditional onramps into professional work may no longer provide the same learning opportunities when AI handles routine analysis. Firms must rethink how they develop junior talent in an environment where the baseline keeps rising.

Professional work under pressure

The study's focus on professional services work is significant. Unlike simple content generation or data entry, these roles demand contextual understanding, stakeholder management, and outputs that executives can act on with confidence. The research suggests that even in these complex domains, AI is reshaping expectations for what junior employees must deliver.

The research team includes Ashish Agarwal, Anitesh Barua, Fangchen Song, and Wen Wen from UT Austin's McCombs School of Business, along with Anu Puvvada, who leads KPMG Studio. Their work examines AI's impact on team collaboration, organizational performance, and the future of work.

These findings were first reported by Harvard Business Review in July 2026.