PwC's Billion-Job Study Reveals AI Is Splitting the Global Workforce Into Winners and Losers — and the Divide Is Accelerating

The most authoritative data portrait yet of artificial intelligence's impact on the global labour market arrived on June 15, 2026, when PwC published its 2026 Global AI Jobs Barometer — an analysis of more than one billion job advertisements across 27 countries and territories spanning six continents. The headline finding is not that AI is eliminating employment wholesale, but that it is cleaving the labour market into two sharply diverging tracks, each with different growth trajectories, wage dynamics, and skill requirements. The report's central concept is a distinction between what PwC calls "professionalised" roles — where AI automates routine tasks while amplifying human judgment and expertise — and "democratised" roles, where AI makes work easier for less-experienced workers to perform, diluting the premium previously attached to specialist knowledge.

The gap between these two tracks is already large and widening quickly. According to PwC, professionalised roles such as radiologists and senior recruiters are seeing twice the growth in available jobs and 42 percent faster salary growth than democratised roles such as IT service managers and medical secretaries. The wage premium for workers possessing AI-specific skills hit 62 percent globally in 2025, up from 57 percent the prior year, and in some consumer-facing sectors the premium exceeds 100 percent. Jobs requiring specific AI capabilities — including machine learning engineering and prompt design — grew by 69 percent from 2024 to 2025, roughly eight times faster than the overall labour market, which expanded by just 9 percent over the same period. For any professional still treating AI fluency as optional, these figures represent an urgent financial argument for reconsideration.

Perhaps the most counterintuitive finding in the Barometer concerns the relationship between AI adoption and headcount. Contrary to the narrative of widespread AI-driven job destruction, PwC found that companies in the most AI-exposed sectors recorded headcount growth of 52 percent relative to 2018 baseline levels, compared to 36 percent for the least AI-exposed companies. The top 20 percent of the most AI-exposed firms achieved average labour productivity growth of 163 percent relative to 2018 — nearly five times higher than the AI-exposed group overall. As PwC's Global Chief AI Officer Joe Atkinson stated in the report release, the companies seeing the greatest returns on AI are using it to amplify human expertise and create entirely new sources of value, not merely to cut costs. The implication for corporate strategy is significant: firms that deploy AI primarily as a headcount-reduction tool may be choosing the slower growth track.

At the entry level, the Barometer's findings are both promising and sobering. An analysis of 2.4 million entry-level job postings in the United States found that positions most exposed to AI are now seven times more likely to require skills traditionally associated with more senior staff — leadership, creativity, and interpersonal communication. Demand for these upgraded entry-level roles has increased by 35 percent since 2019, even as vacancies for conventional entry-level jobs have fallen by 10 percent over the same period. The Federal Reserve Bank of New York separately reported that the unemployment rate for recent college graduates reached 5.6 percent in 2026, above the national average, consistent with a structural hollowing-out of the apprenticeship-style entry-level roles through which previous generations built foundational competencies. A Stanford study of ADP payroll records, cited by SaaSUltra, found that early-career workers aged 22 to 25 in the most AI-exposed jobs saw a 13 percent relative drop in employment — concentrated in roles where AI automates rather than assists.

The strategic read for executives and professionals alike is that the two-track labour market is not a temporary adjustment but an enduring structural feature of AI-era employment. As PwC's report makes clear, the skills needed for the most AI-exposed jobs are changing more than twice as fast as those for the least exposed roles — compressing the time window available for individuals to adapt and organisations to reskill. Companies that use AI primarily to chase efficiency will find themselves on the slower-growth democratised track, while those that use it to unlock new revenue and enter new markets will occupy the professionalised tier where wages and headcount are rising in tandem. The defining career question of the next decade is not whether AI will take your job — it is whether the job AI leaves you with commands a 62 percent wage premium or a 16 percent one.