PwC’s AI Jobs Barometer 2026 found that most-AI-exposed companies grew headcount 52% over the period measured, against 36% for the least-exposed. On its own, that number would settle the AI-jobs argument in favor of growth — and a lot of commentators will stop reading right there.
PwC’s 2026 Global AI Jobs Barometer doesn’t stop there. And what it does next changes the entire question.
The AI Jobs Barometer Number Everyone Will Quote
Released June 15, 2026, the report measured productivity alongside headcount. Most-AI-exposed companies posted 34% productivity growth versus 24% for the least-exposed. AI-related job postings are growing 69% against 9% for the overall job market. The AI skills wage premium sits at 62%, up from 57% a year earlier.
Read in isolation, this is the counter-narrative post AI commentators have been waiting for: AI is creating jobs, not destroying them.
It is also the least interesting part of the report.
The Sorting Mechanism Hiding Inside the Headline
Inside the broader AI-exposed category, PwC identified a “super-star” tier, the top 20% of AI-exposed companies. That tier posted 163% productivity growth. Not double the broader average. Nearly five times it.
The gains are not evenly spread across companies that adopted AI. They are concentrating, hard, into a small group that is pulling away from everyone else, including other companies that also adopted AI.
This is the part the “AI creates jobs” headline misses entirely. AI is not a rising tide. It is a sorting mechanism, and most organizations have no idea which side of the sort they are currently on.
The Sort Is Already Reaching Entry-Level Roles
The sorting shows up earliest where it is least expected: entry-level hiring.
AI-exposed entry-level roles are now 7x more likely to require leadership or judgment skills than they were before. “Seniorised” entry-level postings, roles asking 22-year-olds for what used to be management-track judgment, are up 35% since 2019. Ordinary entry-level postings, the kind that asked for a degree and a pulse, are down 10% over the same period.
We have written about the class of 2026 navigating exactly this shift. The PwC data confirms it at scale: the entry-level job market did not shrink. It got pickier about what an entry-level person needs to already bring.
PwC’s Global Workforce Leader, Pete Brown, put it plainly: AI is removing some of the routine work that once acted as an apprenticeship for early-career workers, while pulling judgment and leadership skills earlier into the career ladder than ever before. (EME Outlook, June 2026)
What Determines Which Side of the Sort You’re On
This is where the report stops being a labor-market story and starts being a leadership story.
The 163% super-star tier is not explained by which AI tools a company bought. Every company in the AI-exposed category has access to roughly the same tools. The super-star companies are the ones whose people brought judgment to those tools, the same TALENT ingredient we have tracked through the software engineering surge and the entry-level data all year.
But TALENT alone does not explain the gap either. Judgment has to be deployed, recognized, and rewarded by someone with the authority to do it. That is LEADERSHIP AT ALL LEVELS: the capacity, at every level of an organization, to recognize where AI created new judgment-shaped roles and to fill them deliberately rather than by accident.
Companies that did both, built the talent and led the deployment of it, are the super-star 20%. Companies that bought the tools and waited to see what happened are the other 80%, growing, but at a fifth of the rate.
What “Building the Talent” Actually Means Here
This is not an argument for hiring more people or buying more AI licenses. PwC’s super-star tier did not win by outspending the other 80%. They won by deciding, early, which roles in their organization needed judgment layered on top of AI output — and then filling those roles deliberately instead of waiting to see which ones turned out to matter.
That decision looks different at every level. At the entry-level, it looks like hiring for the capacity to supervise an AI system’s output, not just operate it — which is exactly what the “seniorised” postings data is measuring. At the management level, it looks like recognizing that a 34% productivity gain and a 163% productivity gain come from the same AI tools, deployed by people who were trusted to exercise judgment rather than simply execute. At the leadership level, it looks like building the organizational habit of asking, every time AI changes a workflow, who in this building now needs more authority, not less.
The 80% are not behind because they lack access. They are behind because nobody made that series of decisions on purpose.
The Reframe
AI does not just create jobs. It is not destroying them either. It is a sorting mechanism, and the sorting criteria are the same five ingredients that determine whether any organization is distinct or extinct in this disruption: IDEAS, SPEED, TALENT, DISTINCTION, and LEADERSHIP AT ALL LEVELS.
The headline number tells you AI is net-positive for employment. It does not tell you whether your organization is in the 20% capturing five times the gain, or the 80% watching from the average. Those are very different positions to be in twelve months from now.
Those prepared need not fear the forces at work.
Where Does Your Organization Stand?
The Kryptonite Scorecard™ measures your organization’s readiness across all five ingredients. Thirty behaviors. Fifteen minutes. A personalized action plan showing exactly where you are prepared and where you are vulnerable.
Take it at realmikeevans.com/scorecard.
Distinct or Extinct: Future-Proofing People and Organizations in the Age of AI is available now on Amazon.
Source: PwC, “2026 Global AI Jobs Barometer,” pwc.com, June 15, 2026.