AI and Jobs series: How entry level jobs are changing in an AI era

Entry-level jobs provide a special window into the wider changes we’ll see to job design and skill development. In this episode of Meet the Leader, Peter Brown of PwC explains how entry-level roles are transforming and breaks down new research on what employers really want from entry-level workers now. He also shares how these roles can continue to serve as training grounds for career skills like collaboration, critical thinking, and relationship building and why a thoughtful redesign of these roles will be key to future leadership development. Top names from our Chief People Officer community also share the advice new grads need to know in a fast-changing job market.

TL;DR

A ~22-minute episode of the World Economic Forum’s Meet the Leader podcast, hosted by Linda Lacina, published 22 July 2026 — the third and final instalment of her AI and Jobs series. The main interview is Peter Brown, global workforce leader at PwC, talking to WEF’s Kateryna Gordiychuk at Summer Davos (the Annual Meeting of the New Champions) in Tianjin, China in June; it is built around a WEF–PwC report on AI and entry-level skills. The episode closes with advice from WEF’s Chief People Officer community.

1. Why entry-level is the diagnostic case, not a special case. Brown declines the framing that entry-level is a separate category: “I don’t actually think of them as a separate category in relation to AI. AI is changing the nature of work at all levels, whether you’re entry level or somebody approaching retirement.” What makes entry-level a useful window is composition: “in those entry-level jobs, much of what the tasks they’ve done traditionally can be automated — but what can’t be automated is actually the skills that are in huge demand.”
2. The headline reframe: this is a productivity-and-skills story, not a jobs story. From PwC’s AI Jobs Barometer: “in those sectors most exposed to AI… the organisations that are getting ahead are seeing three times growth in productivity” — and, crucially, “that’s where the work has fundamentally been redesigned.” His account of why the laggards fail is the same bolt-on diagnosis the wiki tracks elsewhere: “one thing they’re not doing is just layering AI on, bolting AI onto existing ways of working. Some organisations have done that in the chase of efficiencies and you might get some incremental efficiencies. You don’t see the long-term value.”
3. Demand for human skills is growing twice as fast in AI-exposed sectors. “Two times growth rate in the AI-exposed sector — they’re human skills: judgment, problem solving, critical thinking, relationship building.” He puts the resulting question directly: “that’s the fundamental question for leaders — how do you create that experience for entry-level workers to create those skills, which traditionally have been developed through, if you like, the apprenticeship of doing the work and sitting in a room and watching and listening? How do you recreate those experiences… such that in five, seven, ten years’ time we’ve got leaders coming through the pipeline?”
4. The sharpest number in the episode — the seniorization gap. “We know from our research that on entry-level skills, employers are asking for seven times greater what we would call seniorized skills — skills that historically would have developed in the first three to four years. They’re asking for those up front.” His warning about the consequence: “It’s not going to happen by osmosis. You’ve got to be intentional about how you do that.”
5. The report finding Brown says surprised him — some employers are hiring more juniors. “Dare I say sometimes those headlines aren’t backed up by evidence and data… my positive message was that in many cases they’re actually increasing their hiring at that entry level.” One organisation in the report is hiring 25% more, on the explicit rationale “if we don’t invest in the start of the pipeline, how are we going to have leaders in five, ten years’ time?” A second case is a large technology company still developing new coders in “fundamental coding experience before they’re sort of letting them loose into the AI world… because they want coders who have got that human judgment — because at the end of the day, it’s the humans that will be making the decisions, not the technology.”
6. Skills over degrees, and the catch-22 it creates. From the report: 81% of surveyed employers say they prioritise applied experience; 34% prioritise degrees. Brown accepts the trap this sets for graduates — “employers are seeking entry-level workers with applied experience, and yet the paths to get that experience in order to get through the door in the first place can appear to be quite challenging” — and his answer is plural routes (university for some, apprenticeships for others) plus deliberate coordination between business, educators and policymakers. On PwC’s own practice: entry routes cut from “at least 17, 18 ways you could join” to about five, hiring “much more now for potential and skills as opposed to sort of badges and career titles.”
7. His two named misconceptions. First, “that AI automatically means job loss — for me it’s actually around the productivity and the skill story and much more around work design.” Second, “that AI automatically means that human skills become less important, and actually we’re seeing the reverse — as AI is getting more advanced, the requirement for more advanced human skills is increasing.” He pairs this with an urgency point: “when people come and talk about the future of work — well, that’s now.”
8. His durable formulation on job security. “You can protect people. You can’t protect jobs. Jobs will change, they’ll evolve… the difference in the current era is the speed of that change.” His advice to young people is a three-part composite: domain expertise, AI fluency (“take any opportunity to work with AI, to experiment with it”), and deliberate accumulation of human skills through any context where you lead, organise, have difficult conversations, manage conflict or build relationships.
9. The Chief People Officer segment. Carina Cortez (Cornerstone OnDemand): “focus on the outcome or the wisdom versus the task… the risk is if someone does not focus on outcome and is really task-based, that role is going to be obsolete.” Her second point is personal rather than strategic — “if you’re scared, do it anyway” — and she models it by taking a vibe-coding class herself “because I want to understand what we’re asking other people to do.” Maria Flynn (CEO, Jobs for the Future): “where you start is not going to be where you finish… that first job is likely not going to be your dream job,” with the emphasis on extracting identifiable skills and preferences from whatever the first role is.

What was actually ingested

The full auto-generated (ASR) English caption track (615 segments, consistent with duration: 21:55 / length_seconds: 1315). No chapter markers. Speaker turns are not labelled; the episode has four distinct voices (host Lacina, interviewer Gordiychuk, Brown, then Cortez and Flynn), and attributions above were reconstructed from the host’s introductions, which are explicit at each handover. Show open, close and podcast-promotion segments are excluded from the substantive summary. The underlying WEF–PwC report on AI and entry-level skills is not ingested — every statistic here is Brown’s verbal characterisation of it.

Dynamic-capabilities tagging

• digital-transforming/redesigning-internal-structures — the episode’s central prescription is structural rather than technological: redesign the entry-level role so that the apprenticeship function survives the automation of the tasks that used to carry it. PwC’s own restructuring of entry routes (17–18 down to about five) and its shift to hiring for potential and skills rather than degrees and titles are worked instances, as is the report’s case of an employer raising entry-level hiring 25% on a leadership-pipeline rationale.
• contextual/internal-barriers — the seniorization gap is named as a barrier that will not resolve itself: employers now want, up front, skills that used to accrete over three to four years, and “it’s not going to happen by osmosis.” Brown treats the absence of intentional experience design — not AI capability — as the thing standing between current practice and a functioning leadership pipeline.

Linked entities and concepts

• PwC — Brown’s employer and co-author of the underlying report; also the source of the AI Jobs Barometer and the 55,000-worker Hopes and Fears survey cited. Updated in this ingest.
• McKinsey — the same entry-level-rung problem from a shared-services operations vantage; see this source’s relationships:.
• Emerson, Kropp et al. — the same reshape-over-replace reading of the survey evidence.
• Brynjolfsson — the typed contradicts edge above: employer-survey optimism against ADP payroll data showing a growing entry-level employment decline in the most exposed occupations.
• ai-employment-effects — the 3× productivity, 2× human-skills-demand, 7× seniorization and 81%/34% figures, and the increasing-entry-level-hiring counter-cases.
• ai-deskilling — the apprenticeship mechanism: the tasks that carried skill formation are the ones being automated, and the large technology company deliberately teaching fundamental coding first is a countermeasure.
• durable-skills — judgment, problem solving, critical thinking and relationship building as the skills whose demand is growing fastest; AI fluency as the third leg alongside domain expertise and human skills.
• micro-productivity-trap — “bolting AI onto existing ways of working… you don’t see the long-term value,” with the 3× productivity gain conditioned on fundamental work redesign.
• dynamic-capabilities — the two cells tagged above.

Dangling (single-source mention, deferred per author-entity promotion): the World Economic Forum as a publishing channel (first appearance; promote on a second WEF source), Peter Brown, Linda Lacina, Kateryna Gordiychuk, Carina Cortez, Maria Flynn, Jobs for the Future, Cornerstone OnDemand.

Source quality note

Auto-generated transcript; proper nouns corrected at acquire time (PwC, Linda Lacina, Kateryna Gordiychuk, Tianjin, Carina Cortez, Cornerstone OnDemand — see the raw file’s notes:).

Every quantitative claim is second-hand and unverified here. The 3× productivity, 2× human-skills demand, 7× seniorization and 81%/34% figures are Brown’s verbal summary of a report co-authored by his own firm and published with the WEF; neither the report nor PwC’s AI Jobs Barometer has been ingested, so sampling frames, definitions (“seniorized skills,” “AI-exposed sector”) and the direction of causality are all unexamined. The two encouraging cases (25% more entry-level hiring; a technology company teaching fundamentals first) are anonymised and unquantified beyond those phrases.

Note the interest structure, which cuts both ways: PwC sells workforce-transformation advisory, so a “redesign work, don’t cut heads” conclusion is commercially convenient — but PwC is also a very large employer of entry-level graduates, and Brown speaks partly as a practitioner describing his own firm’s hiring changes. Brown is careful to flag his own priors (“I’m the father of a 20-year-old”). The contradicts edge to Brynjolfsson is the most useful check the wiki can currently apply to the optimistic reading: employer-stated intent is a weaker evidence class than payroll records.