If AI does the junior work, who will become tomorrow’s seniors?

Think about the work many professionals did at the start of their careers: research, first analyses, data cleaning, summaries, presentations and drafts reviewed by someone more experienced. A growing share can now be done much faster with AI.

From a productivity perspective, the logic is understandable. If an experienced employee with AI can complete in two hours what previously occupied a junior for much longer, organisations will naturally question how many people they need at the bottom of the pyramid.

But those hours were never only about producing an analysis or presentation. They were also how someone learned to analyse, present and make decisions.

Repetition builds judgement. By analysing cases, patterns start to emerge. Drafting and receiving feedback from your manager sharpens your skills. Observing experienced colleagues reveals questions you had not considered. And mistakes become valuable learning opportunities while the consequences are still manageable.

Most organisations never classified these moments as Learning & Development. They simply happened through work.

That is why this shift goes beyond which tasks AI can automate. AI may also be changing how expertise is built.

Stanford’s Digital Economy Lab analysed payroll data covering millions of US workers through June 2026. It found no evidence of widespread economy-wide job displacement, but did find a clear difference at the start of careers.

Among workers aged 22 to 25 in AI-exposed occupations, employment stood 19% below where it would have been if it had kept pace with less-exposed peers. More experienced workers did not show a comparable gap. The difference was mainly driven by reduced hiring of younger workers rather than increased dismissals. (Stanford Digital Economy Lab)

That does not mean AI caused a 19% decline. Stanford describes the findings as early, descriptive signals rather than causal proof. But they suggest that early-career employment may be affected differently as AI becomes embedded in organisations.

PwC’s 2026 AI Jobs Barometer points to another part of the shift. Junior roles with high AI exposure are seven times more likely than low-exposure junior roles to require skills traditionally associated with seniority, including leadership and strategic thinking. Skills in highly AI-exposed jobs are also changing more than twice as fast. (PwC)

That suggests the junior role may not simply disappear. It may become more senior.

A junior marketer can produce a first draft much faster, but increasingly adds value by judging its quality, relevance and reliability. A junior analyst can produce an initial analysis faster, but still needs to recognise when the numbers tell the wrong story.

AI makes execution easier. Judgement does not automatically appear in its place.

This is where the AI discussion becomes an L&D discussion.

Learning & Development has often been organised around the job rather than inside it. Employees work, while alongside that work they follow training, join programmes or create development plans.

Yet much professional development never happened through formal interventions. It happened because people were doing the work.

If AI removes some of the work through which people traditionally learned, organisations cannot assume development will continue in the same way.

The key question for L&D therefore becomes less about which training to offer and more about which experiences people need to become genuinely good at their jobs.

That may mean exposing junior employees to decision-making earlier, providing more structured mentoring, involving them sooner in complex projects or using AI-generated time savings for customer exposure and cross-functional work.

PwC reaches a similar conclusion, arguing that organisations need to redesign early-career pathways, onboarding, mentorship and training so junior employees can build higher-level skills earlier. (PwC)

This brings L&D much closer to workforce planning. When learning no longer happens automatically through repetition, development has to be designed more deliberately.

There is an important upside. AI can remove learning opportunities, but it can also create new ones.

People early in their careers now have access to explanations, feedback and knowledge that previously depended heavily on an experienced colleague. They can test arguments, explore alternatives and work through problems much faster.

Used well, AI could shorten parts of the learning curve considerably.

The real question is what organisations do with the time that is freed up. If five hours saved simply become five more hours of AI-assisted output, productivity may increase without much additional development.

If those hours allow someone to join a complex project, observe a senior colleague, speak to customers or take responsibility earlier, AI can become an accelerator of development rather than a shortcut around it.

That is why measuring AI success only in hours saved or headcount reduced is too narrow. A better question is: what can people now learn and do that they could not do before?

There is also a longer-term workforce issue.

An organisation may need fewer juniors because AI makes its existing workforce more productive. In the short term, that can make perfect sense. A few years later, however, that same organisation needs experienced people. Where will they come from?

One answer is the external labour market. Instead of developing juniors internally, recruit seniors when you need them. That works as long as enough other organisations continue investing in junior talent. If many companies make the same calculation, everyone eventually competes for the same smaller pool of experienced professionals.

Companies then effectively outsource their talent pipeline to organisations that did hire people early, teach them, tolerate their mistakes and gradually increase their responsibility.

There is an important difference between automating junior tasks and eliminating junior development. The first may become unavoidable. The second is a strategic choice.

AI also makes years of experience a weaker proxy for expertise.

Two people can spend three years in the same role and accumulate very different experiences. One may use AI mainly to execute existing work faster. Another may use the capacity created by AI to join complex projects, work across teams and take responsibility earlier.

On paper, they have the same tenure. Their readiness for the next step may be very different.

Career development may therefore need to become less about time in role and more about accumulated experiences, developed capabilities and potential for what comes next.

This also brings L&D, internal mobility, career conversations, mentoring and succession planning closer together. They are different parts of the same question: How do we create the people our organisation will need next?

Answering that requires understanding employees beyond their job titles and current skills. Organisations need to know what people are good at, what motivates them, where they want to go and what they still need to learn.

That is also why this topic is relevant to our work at myCareerCompanion. Not because AI suddenly requires a new HR tool, but because less predictable career paths make it more important to understand people beyond the role they occupy today.

AI can make organisations more productive and help people at the beginning of their careers contribute at a higher level sooner. That is a real opportunity.

But junior work has always done two jobs. It produced output, and it produced more experienced professionals.

AI may take over part of the first. Organisations will have to become much more deliberate about the second.

The companies that get this right may not be those that automate the most junior work. They may be those that redesign learning, career development and early-career experiences fast enough to benefit from AI without weakening their future talent pipeline.

Because AI can do more of the junior work.
Someone still has to become the senior.