For the last two years, the conversation about AI and employment has largely been framed as a single question: is AI going to take jobs or not? New global research suggests that framing was always too simple. AI isn't uniformly replacing or protecting roles. It's splitting the labour market into two genuinely different tracks, and which track an organization or a role ends up on depends less on the industry than most leaders assume.
A Billion Job Postings, One Clear Pattern
PwC's 2026 Global AI Jobs Barometer analyzed more than one billion job advertisements across 27 countries, combining large-scale labour market data with company financial and occupational task information.
The headline finding is what PwC calls a two-track labour market. On one track sit "professionalised" roles, positions where AI automates routine tasks, freeing up human judgment and expertise to become more valuable rather than less.
Professionalised Roles
AI automates routine tasks while human judgment and expertise become more valuable. Recruiters, radiologists, and similar roles fall into this category.
Democratised Roles
AI makes the underlying work easier to perform, reducing the specialized expertise needed to do it.
The gap between these two tracks is widening quickly. Professionalised roles are seeing roughly twice the growth in available positions compared to democratised ones.
The most AI-exposed companies, the top 20 percent by PwC's measure, achieved average labour productivity growth of 163 percent relative to 2018, nearly five times higher than the broader group of AI-exposed companies.
The Wage Premium Is Growing, Not Shrinking
One of the clearest signals in this data is the widening pay gap tied to AI skills specifically.
In the UK, AI skills now deliver a higher wage return than a master's degree, a 23 percent premium compared to 13 percent for an advanced degree.
This isn't just a technology story. It's a signal about where value is actually being created inside organizations.
A Harvard Business School study analyzing nearly all US job postings from 2019 through early 2025 found that openings for routine, automation-prone roles declined 13 percent after generative AI tools became widely available, while demand for analytical, technical, and creative roles grew 20 percent over the same period.
The labour market isn't shrinking overall. It's reorganizing around a different definition of what counts as valuable human work.
Entry-Level Work Is Where the Split Is Sharpest
The most consequential part of this year's PwC research is its targeted look at entry-level roles specifically, and the picture there is more complicated than the aggregate numbers suggest.
In highly AI-exposed occupations, the skill requirements of early-career jobs are changing rapidly.
Fewer Routine Tasks
Tasks that once served as the on-ramp for junior professionals are increasingly being performed by AI.
Higher Expectations
Remaining entry-level tasks increasingly require judgment and expertise that junior candidates have not yet developed.
A Pipeline Challenge
Roles designed to build experience can shrink at the same time that experienced judgment becomes more valuable.
Future Expertise
Organizations need to think deliberately about how junior talent develops the judgment AI cannot replicate.
Organizations that don't think deliberately about how junior talent develops the judgment AI can't replicate risk hollowing out their own future pipeline of senior expertise, even as they benefit from AI's productivity gains in the near term.
New Roles Are Emerging, But Unevenly
The World Economic Forum's Future of Jobs Report projects that by 2030, AI-driven disruption will affect 22 percent of all jobs globally, displacing 92 million existing positions while creating 170 million new ones.
The fastest growth concentrates in technology, data, and AI roles specifically, but meaningful growth is also expected in healthcare, education, and the green economy.
Entirely new job categories are emerging as a direct consequence: LLM fine-tuning specialists, retrieval-augmented generation pipeline engineers, AI ethics auditors, and prompt engineers are all roles that didn't exist five years ago.
The World Economic Forum also identified environmental stewardship as a top-growing skill for the first time this year, reflecting a crossover between AI-driven sustainability reporting and climate-focused work that most workforce planning hasn't caught up to yet.
The Leadership Gap Is the Real Bottleneck
Despite the scale of this shift, most organizations aren't managing it well.
Deloitte's 2026 Global Human Capital Trends report found that only 6 percent of leaders believe they're making real progress designing how humans and AI should actually work together.
That's a striking number given how much capital and attention AI has absorbed across nearly every industry.
The technology is moving faster than the organizational thinking required to deploy it well, and the World Economic Forum estimates that employers expect 39 percent of workers' core skills to change by 2030.
What This Means for Canadian Organizations
For business and HR leaders, the practical implication isn't a binary choice about whether to adopt AI. Nearly everyone is adopting it in some form already.
The more useful question is which track a given role, team, or function is on, and whether that's the track the organization actually wants it on.
When AI Elevates Human Work
Roles where AI is meant to free up human judgment and expertise for higher-value work need deliberate investment in the humans doing that work: training, development, and a genuine plan for how junior talent gains the experience that judgment requires.
When AI Makes Work Easier
Roles where AI genuinely makes the work easier and more accessible need a different kind of planning, focused on how the organization redeploys the capacity that gets freed up.
The organizations pulling ahead in this data aren't simply the ones using AI the most.
They're the ones being deliberate about where AI substitutes for human work and where it's meant to elevate it, and building their hiring, training, and workforce planning around that distinction rather than treating AI adoption as a single undifferentiated strategy.