This article explores whether workplace AI is quietly shrinking the first rung of the career ladder, and what leaders should do to protect talent pipelines while still capturing productivity gains.
After reading this article you will understand what changed in March 2026, what the evidence actually shows about entry level hiring, and how to redesign early career roles without weakening future capability.
What changed in March 2026
This month brought a meaningful capability shift. OpenAI says GPT 5.4 is designed for computer use workloads and can operate across applications, with stronger benchmark performance in multi step tasks. That does not mean enterprise automation is suddenly solved, but it does mean the ceiling has moved.
The second shift is where AI is appearing. OpenAI’s ChatGPT for Excel places AI directly inside the workbook and reports improved performance on internal financial benchmarks. That matters because spreadsheet work has long been a core entry level proving ground. If models can now draft, structure, and update work inside the file itself, leaders need to decide which parts of that work still build capability and which have become commodity.
Microsoft’s March announcements point to a different issue. The company is not only adding agent features, it is reframing the problem around trust, control, and observability. As outlined in Microsoft’s March 2026 announcement, Agent 365 is designed to give organisations visibility and control over agents at scale.
Its Copilot updates also highlight expanded skill inferencing based on work activity, effectively turning workplace behaviour into inferred workforce intelligence. That creates value, but it also introduces privacy and employee monitoring questions that most organisations have not fully addressed.
What the evidence says, and what it does not
The strongest mistake leaders can make here is to confuse a survey signal with settled labour market fact. Resume.org’s March press release says 21 percent of surveyed companies have stopped hiring entry level workers because of AI. That is useful as a signal of employer perception, but it is not economy wide proof of behaviour.
The ECB’s March 2026 analysis provides a more grounded European perspective. It finds that firms making significant use of AI are, on average, more likely to hire than fire, suggesting early adoption is often complementary to workforce expansion rather than substitution.
Anthropic’s labour market research reinforces that caution. It reports limited evidence of employment impact to date, while explicitly stating that findings should be revisited as adoption deepens.
The New York City data adds another dimension. Entry level postings have fallen sharply over several years, and internship opportunities have declined at a similar rate. That trend predates widespread enterprise AI deployment, which suggests macroeconomic conditions remain a major driver, even if AI is now influencing how firms respond.
The UK context sharpens the stakes further. According to the Office for National Statistics, 957,000 young people aged 16 to 24 were not in education, employment, or training in late 2025, highlighting how fragile entry routes already are.
Why the first rung matters more than leaders think
Entry level roles are often misunderstood. They are not just about completing routine tasks, they are where people learn how work actually operates in practice. Junior staff develop judgement through repetition, correction, and exposure to real situations, gradually building an understanding of how decisions are made and why certain outputs matter more than others.
Over time, they learn to interpret ambiguous requests, validate outputs, and navigate processes that rarely behave as neatly as they appear on paper. That experience is difficult to replicate through training alone because it depends on context, feedback, and the accumulation of small decisions made under supervision.
If organisations remove the repetitive element without redesigning the developmental element, they risk weakening the long term pipeline. A workforce cannot continuously hire experienced talent without also developing it, which means entry level roles must still provide a pathway into higher responsibility work.
This makes the issue less about hiring volumes and more about job design. Leaders need to separate tasks from roles and roles from pathways, deciding deliberately which tasks should disappear and which should remain because they build capability.
Which junior work is most exposed
The tasks most exposed are already visible. AI is now embedded in drafting, summarising, spreadsheet modelling, document synthesis, and basic research workflows, all of which have traditionally formed the core of entry level knowledge work across functions.
These activities have historically provided a structured way for junior staff to build speed, confidence, and familiarity with tools and data. If those tasks are compressed or removed entirely, the role itself does not disappear, but its purpose changes.
The developmental value therefore has to shift upwards. Junior staff should spend less time producing outputs and more time evaluating them, testing assumptions, and understanding how those outputs connect to decisions and outcomes. This requires more deliberate role design rather than simply layering AI on top of existing responsibilities.
A useful test for leaders is whether the redesigned role still builds judgement. If it does not, the pathway is weakening, even if productivity appears to improve in the short term.
What leaders should do now
Start by defining entry level work as a set of tasks and learning outcomes rather than a job title. Identify which tasks build long term capability and which do not.
Then map which tasks are now automatable. Drafting, summarising, and repetitive analysis are obvious candidates. High impact decisions and sensitive workflows should remain supervised.
The next step is redesign. Entry level roles should include structured review, escalation, and accountability. AI should assist, not replace the learning process.
Finally, measure whether pathways are weakening. Track hiring rates, internships, conversion, and early performance indicators. Most organisations measure headcount. Few measure whether capability is still being built.
Governance is now part of workforce strategy
This is not only a productivity issue. It is also a governance issue that cuts across hiring, role design, and day to day decision making.
As AI becomes more agentic, organisations need visibility, control, and accountability over how work is done and who or what is making decisions. Regulatory signals from the EU, UK, and US all point in the same direction, particularly around transparency, data use, and automated decision making in employment contexts.
Governance therefore cannot sit in a separate compliance function. It has to shape how roles are designed, how data is accessed, and how decisions are made across the organisation, ensuring that productivity gains do not come at the expense of trust, explainability, or legal risk.
Conclusion
The biggest near term workforce risk from AI may not be sudden job losses. It may be the gradual erosion of entry level pathways.
The evidence does not support panic, but it does support action. Hiring trends are mixed, and macro factors remain significant, but capability changes are real.
Leaders therefore need a disciplined response. Protect the learning value of junior work. Automate repetition. Redesign roles. Build governance early.
Because once the first rung disappears, rebuilding it is far harder than automating it ever was.
About the author: Kieran Gilmurray
Kieran is a globally recognized authority on AI, automation, and digital transformation, having authored multiple influential books and hundreds of articles that have earned him prestigious accolades, including being named a Top 50 Global Thought Leader and Influencer on Generative AI in 2024, a Best LinkedIn Influencer for AI and Marketing, Top 50 Global Thought Leaders and Influencers on Manufacturing 2024, Top 14 people to follow in data and one of the World’s Top 200 Business and Technology Innovators.
CLICK HERE TO SCHEDULE AN ANALYST CHAT
Links:
Originally posted in the IRPA AI Network — Announcements & Updates