The integration of generative artificial intelligence (AI) in the workplace is reshaping how junior employees acquire essential skills, raising concerns among business leaders about the long-term development of professional expertise. Traditionally, entry-level roles have combined task completion with repeated practice that helps build judgement and critical thinking. However, as AI assumes responsibility for routine functions, the conventional pathway for skill development is being disrupted.

James Tucker, global lead for corporate finance and strategy at Boston Consulting Group (BCG), observes that finance teams are shrinking as AI takes over repetitive tasks, leaving greater emphasis on judgement and quality control. While this improves efficiency, it complicates the mentoring and training process for new hires, who historically gained expertise through numerous iterations of basic assignments.

A recent BCG study involving 70 senior leaders found that about half are witnessing a decline in skills among their teams, with 53% reporting slower advancement of junior talent. Supporting this trend, a Harvard working paper analyzing over 280,000 U.S. firms identified a decrease in junior hiring at companies adopting generative AI technologies. The reduction is largely due to slowed recruitment rather than layoffs, primarily affecting roles most exposed to AI. Similar patterns are evident in the United Kingdom, where government data shows weakening entry-level hiring in knowledge-intensive sectors such as accountancy, software engineering, and data analysis amid broader labor market slowdowns.

Jack Kennedy, senior economist at Indeed, cautions that diminished hiring opportunities pose significant challenges for graduates and younger workers striving to gain initial professional experience. He notes that early career setbacks can have lasting effects over a person’s working life. This concern is echoed in Deloitte’s latest UK CFO survey, where 47% of respondents anticipate a decline in graduate recruitment over the coming year due to AI. Meanwhile, analysis by PwC reveals that entry-level roles highly susceptible to AI intervention now often require skills traditionally associated with more senior employees, heightening expectations for new hires.

As AI takes over more execution-level work, employers increasingly prioritize human skills like interpretation, verification, and judgement. Sharon Steiner, chief human resources officer at Fiverr, highlights a shift in hiring focus: candidates may not need to be top coders but must demonstrate strategic business insight, a valuable capacity when AI can instantly generate technical outputs. Moreover, she emphasizes that skill development is not confined to formal employment; side projects and independent initiatives also contribute to building expertise.

Research by Microsoft involving 319 knowledge workers indicates that greater confidence in generative AI corresponds with lower self-reported critical thinking, though the nature of critical thinking evolves toward oversight and validation rather than disappearance. In finance, for example, AI can produce routine reconciliations quickly, but signing off such work requires experience to detect errors or misleading results. Martino Cadoni, CFO at DeepL, underscores the necessity of human scrutiny and ongoing training to maintain a critical mindset when collaborating with AI systems.

To address these challenges, some organizations are redesigning training programs to blend traditional exercises with AI-assisted tasks, fostering judgement and scepticism alongside technical skills. Deloitte UK plans to restructure parts of its graduate audit training from September 2026, incorporating simulations that juxtapose conventional methods with AI-generated outputs. Anthony Salcito, senior vice-president at Coursera, advocates for onboarding processes that combine role-specific learning, practical experience, and clear guidance on when human intervention is crucial, emphasizing the need for safe environments to experiment with AI tools without high stakes.

Other companies, like Shell, are taking deliberate steps to preserve learning opportunities by requiring juniors to frame problems and validate assumptions before turning to AI, followed by peer review of key results. BCG recommends structured debates and AI-free problem-solving sessions as part of developing professional judgement.

While AI clearly enhances short-term productivity by accelerating analysis and reducing routine workloads, the longer-term implications for workforce capability are less certain. Smaller cohorts of junior employees could lead to thinner talent pipelines and greater dependence on external hires, increasing the need for senior managers to devote time to coaching and oversight.

Experts caution against viewing AI adoption solely through a lens of labour cost savings. Cadoni stresses that focusing narrowly on immediate productivity gains risks overlooking vital factors such as work quality and the duration required to produce it. The essential challenge for businesses is to find new ways to cultivate expertise and professional judgement, even as the traditional formative experiences associated with routine tasks fade. The success of AI in transforming workplaces will ultimately hinge on how well organizations adapt their training and development models to ensure a skilled and capable workforce for the future.