Data & benchmarks
AI job displacement statistics (2026)
Direct answer
The largest studies of AI and work do not agree on a single number, but they converge on a pattern. The World Economic Forum projects that by 2030 AI and related trends will create 170 million jobs and displace 92 million — a net gain of 78 million and churn in about 22% of today's jobs. Goldman Sachs estimates 300 million full-time jobs are exposed to automation worldwide, the IMF puts AI exposure at about 40% of global employment (60% in advanced economies), and McKinsey expects up to 30% of US work hours to be automatable by 2030. This page collects those figures with a link to each primary source.
Key takeaways
- Displacement and creation happen together: the WEF projects 92 million jobs displaced and 170 million created by 2030, a net gain of 78 million.
- Exposure is broad but is not the same as loss: the IMF estimates ~40% of global jobs are exposed to AI, and Goldman Sachs counts 300 million full-time jobs affected.
- The near-term reality is augmentation more than replacement: most exposed roles have only part of their tasks automatable, and AI-skilled workers earn a 56% wage premium (PwC).
- Reskilling is the decisive variable: the WEF says 59% of workers will need reskilling or upskilling by 2030, and 11% of them may not receive it.
At a glance
Key statistics
The most-cited numbers on AI job displacement, each attributed to its primary source. These are the single-sentence figures most often quoted; the sections that follow put them in context.
- By 2030, AI and related trends are projected to create 170 million jobs and displace 92 million, a net gain of 78 million and churn in about 22% of today's jobs. (World Economic Forum, Future of Jobs Report 2025)
- Generative AI could expose the equivalent of 300 million full-time jobs to automation worldwide. (Goldman Sachs Research, 2023)
- About 40% of global employment is exposed to AI, rising to roughly 60% in advanced economies. (International Monetary Fund, 2024)
- Up to 30% of the hours worked in the US economy could be automated by 2030 as generative AI accelerates adoption. (McKinsey Global Institute, 2023)
- 27% of jobs are in occupations at high risk of automation across OECD countries. (OECD Employment Outlook, 2023)
- Workers with AI skills command a 56% wage premium, and productivity growth has nearly quadrupled in the most AI-exposed industries. (PwC, 2025 Global AI Jobs Barometer)
- 59% of the global workforce will need reskilling or upskilling by 2030, and 11% of them are unlikely to receive it. (World Economic Forum, Future of Jobs Report 2025)
- Today 47% of work tasks are done mainly by humans, 22% mainly by machines, and 30% by a mix — a split employers expect to be roughly even by 2030. (World Economic Forum, Future of Jobs Report 2025)
- 52% of US workers are worried about AI's future use in the workplace, and 32% expect it to mean fewer job opportunities for them. (Pew Research Center, 2025)
Displaced vs created
Jobs displaced vs jobs created
The headline debate is whether AI is a net job destroyer or job creator. The most-cited macro projections point to net creation by 2030, alongside significant churn as some roles shrink while others grow.
The World Economic Forum's Future of Jobs Report 2025, drawing on more than 1,000 employers representing 14 million workers, projects a net increase of 78 million jobs by 2030: 170 million created against 92 million displaced. That net figure hides a larger gross reshaping — roughly 22% of all jobs are expected to change one way or the other. It is worth separating two distinct concepts that headlines often merge. "Displacement" means specific roles disappearing; "exposure" means a job contains tasks that AI could perform. Goldman Sachs Research's widely quoted estimate that 300 million full-time jobs are exposed to generative-AI automation counts the second, not the first.
Early real-world data leans toward creation and transformation rather than wholesale loss. PwC's 2025 Global AI Jobs Barometer, which analysed close to a billion job advertisements across six continents, found that job numbers are still growing even in the occupations most exposed to automation. The pattern that emerges from the macro studies is not "AI destroys jobs" but "AI reshuffles them" — a churn that is manageable in aggregate yet disruptive for the specific workers and roles on the wrong side of it.
Exposure & scale
How much work is exposed to AI
"Exposure" measures how much of a job's work AI could plausibly do. It is the single most useful lens because it captures augmentation and replacement together, before any judgement about which will win out.
| Measure | Figure | Source |
|---|---|---|
| Global employment exposed to AI | ~40% (≈60% in advanced economies) | IMF (2024) |
| US & European occupations exposed to some AI automation | ~two-thirds | Goldman Sachs (2023) |
| Current work tasks AI could substitute | up to ~25% | Goldman Sachs (2023) |
| Jobs in occupations at high risk of automation (OECD) | 27% | OECD Employment Outlook (2023) |
| US workers in the most AI-exposed jobs (2022) | 19% | Pew Research Center (2023) |
The International Monetary Fund's 2024 analysis, Gen-AI: Artificial Intelligence and the Future of Work, found that about 40% of global employment is exposed to AI, rising to roughly 60% in advanced economies where cognitive, office-based work is more common. Crucially, the IMF splits that exposure roughly in half: for about one half of exposed jobs, AI is likely to complement workers and raise productivity, while for the other half it could substitute for tasks and lower labour demand. Goldman Sachs reached a similar order of magnitude from a different angle, estimating that around two-thirds of occupations in the US and Europe are exposed to some degree of AI automation and that the technology could substitute for up to a quarter of current work.
The OECD adds an important caveat about what "high risk" really means. In its 2023 Employment Outlook, 27% of jobs sit in occupations at high risk of automation — but even in those occupations, only an estimated 18-27% of the underlying skills and abilities are highly automatable, so the OECD stresses that at-risk jobs will not simply disappear. Pew Research Center's occupational analysis lands lower still, placing 19% of US workers in the most-exposed jobs as of 2022. Read together, the exposure estimates range widely because they measure different things — but none of the major sources concludes that most exposed jobs vanish outright.
Timelines
Timelines: when the shift happens
Most large studies use 2030 as their horizon. Generative AI has pulled forward the pace of change without collapsing the timeline into a single moment.
McKinsey Global Institute's Generative AI and the future of work in America (2023) quantifies the acceleration precisely: activities accounting for up to 30% of the hours worked across the US economy could be automated by 2030, up from 21.5% in the same model without generative AI. In other words, generative AI added roughly eight percentage points of automatable work and moved the timeline forward, but the bulk of the transition still plays out over the second half of the decade rather than overnight.
McKinsey also frames the human side of that timeline. It expects an additional 12 million occupational transitions in the US by 2030, with an estimated 11.8 million workers currently in shrinking occupations — office support, customer service, and production roles chief among them — likely needing to move into different lines of work. The World Economic Forum's 22% churn figure describes the same phenomenon on a global scale and over the same 2030 horizon. The consistent message across sources is that the disruption is a multi-year structural transition, giving employers and workers a window to reskill rather than a cliff edge.
Most affected
Which roles and industries are most affected
Exposure is highly uneven. Unusually for a wave of automation, this one reaches deep into white-collar and cognitive work, not just routine manual tasks.
| Occupation group | Share of tasks exposed |
|---|---|
| Office & administrative support | 46% |
| Legal | 44% |
| Architecture & engineering | 37% |
| Construction & extraction | 6% |
| Installation, maintenance & repair | 4% |
| Building & grounds cleaning & maintenance | 1% |
Source: Goldman Sachs Research (2023).
Goldman Sachs found office and administrative support the most exposed occupation group, with 46% of its tasks susceptible to automation, followed by legal (44%) and architecture and engineering (37%). At the other end, physical, hands-on trades were barely touched: construction and extraction (6%), installation, maintenance and repair (4%), and building and grounds cleaning (1%). The World Economic Forum's role-level view aligns with this: it names cashiers and ticket clerks, administrative assistants and executive secretaries, and — increasingly — bank tellers, postal clerks, data-entry clerks, and even graphic designers among the fastest-declining roles, while big-data specialists, fintech engineers, and AI and machine-learning specialists top the fastest-growing list.
What makes this wave distinctive is its reach into skilled, well-paid cognitive work. Pew Research Center found that AI exposure rises with education and pay, with occupations such as budget analysts, tax preparers, and web developers sitting in the most-exposed quartile. That is close to the mirror image of earlier automation, which concentrated on routine manual labour. The OECD notes that low- and middle-skilled jobs remain most at risk overall, but the addition of generative AI has extended meaningful exposure into analytical and creative professions that were long assumed to be safe. By industry, PwC's barometer finds the sharpest effects in financial services and software — the sectors adopting AI fastest.
Augmentation vs replacement
Augmentation vs replacement
The near-term evidence points more toward AI changing how work is done than eliminating the worker — with real wage and productivity effects already showing up.
The clearest single indicator of augmentation is the human-machine task split. The World Economic Forum reports that in 2025 about 47% of work tasks are performed mainly by humans, 22% mainly by machines and algorithms, and 30% by a combination of the two — and employers expect that mix to move toward a roughly even split by 2030. The direction of travel is toward collaboration rather than clean substitution, which is why the WEF frames the goal as designing technology to complement human work. The IMF's finding that roughly half of exposed jobs stand to benefit from AI rather than be replaced points the same way.
The labour-market data backs this up. PwC's 2025 barometer found that productivity growth has nearly quadrupled in the industries most exposed to AI, that revenue per employee is growing about three times faster there than in the least-exposed industries, and that workers who bring AI skills earn a 56% wage premium over otherwise-similar peers. Goldman Sachs, meanwhile, estimated AI could lift global GDP by about 7% — roughly $7 trillion — over a decade. These are augmentation signals: value and pay concentrating around people who use AI well. The counterweight comes from the WEF, where 40% of employers still expect to reduce their workforce where AI can automate tasks, and from the IMF's warning that these gains are likely to widen inequality.
Reskilling & sentiment
Reskilling and worker sentiment
Across every major study, the deciding factor is not whether tasks are automatable but whether workers are reskilled in time. Worker sentiment reflects the stakes.
The World Economic Forum quantifies the reskilling challenge in stark terms: 59% of the global workforce will need reskilling or upskilling by 2030, nearly 40% of the skills required on the job will change or become outdated, and 63% of employers already name the skills gap as the single biggest barrier to transformation. The response is broad — some 85% of employers plan to prioritise upskilling their people, and about half plan to move staff from declining roles into growing ones — but the gap is real: the WEF estimates that 11% of workers who need reskilling are unlikely to receive it, which equates to more than 120 million people at medium-term risk. Regionally, North American employers expect 67% of their workforce to need training, the highest of any region.
Worker sentiment tracks that uncertainty. Pew Research Center's 2025 survey found that 52% of US workers are worried about AI's future use in the workplace and 32% expect it to reduce their long-term job opportunities, against 36% who feel hopeful. The OECD found a similar mood internationally, with three in five workers concerned about losing their job to AI within a decade — yet two-thirds of those already working alongside AI said it had made their work less monotonous or dangerous. The lived experience of AI, in other words, tends to be less alarming than the anticipation of it, which is one reason reskilling and hands-on exposure matter so much to how the transition actually lands.
Actigy perspective
The operational read: AI plus human operations
Our reading of this evidence, as an operations provider, is that the near-term winner is neither pure automation nor pure headcount but the combination: AI absorbs routine, high-volume work while skilled people handle exceptions, judgement, and quality. That is the model behind Actigy's AI plus human operations approach and AI outsourcing services, and it mirrors what the data shows — augmentation ahead of replacement, with reskilling as the deciding variable.
Methodology
Sources & methodology
Every statistic on this page is drawn from a published third-party report and linked to its source. Actigy does not claim these figures as its own data. The estimates come from different methodologies — employer surveys, occupational task modelling, and analysis of job-advertisement data — and use different definitions of "exposure," "risk," and "displacement," which is why the numbers vary; each figure should be read with its source's method and date in mind. Figures reflect the most recent editions available as of July 2026.
- World Economic Forum — Future of Jobs Report 2025 (2025) — jobs created and displaced by 2030, 22% churn, task split, reskilling needs
- World Economic Forum — The fastest-growing and declining jobs (2025) — fastest-growing and fastest-declining roles
- Goldman Sachs Research — Generative AI could raise global GDP by 7% (2023) — 300 million jobs exposed, occupation exposure shares, GDP impact
- McKinsey Global Institute — Generative AI and the future of work in America (2023) — US work hours automatable by 2030, occupational transitions
- International Monetary Fund — Gen-AI: Artificial Intelligence and the Future of Work (2024) — 40% global / 60% advanced-economy AI exposure
- PwC — 2025 Global AI Jobs Barometer (2025) — 56% AI wage premium, productivity growth, jobs rising in automatable roles
- OECD — OECD Employment Outlook 2023 (2023) — 27% of jobs at high risk of automation, worker concern
- Pew Research Center — Which US workers are more exposed to AI on their jobs? (2023) — 19% in most-exposed jobs, exposure by occupation
- Pew Research Center — US Workers Are More Worried Than Hopeful About Future AI Use in the Workplace (2025) — worker sentiment on AI
Found this useful? You are welcome to cite or link to this page. Last updated July 2026.
FAQ
AI job displacement: FAQ
How many jobs will AI displace by 2030?
The World Economic Forum's Future of Jobs Report 2025 projects that 92 million jobs will be displaced by 2030 while 170 million new roles are created, a net increase of 78 million and total churn equal to about 22% of today's jobs. Separately, Goldman Sachs estimated in 2023 that generative AI could expose the equivalent of 300 million full-time jobs to automation worldwide, though exposure is not the same as loss.
Will AI create more jobs than it destroys?
On current projections, yes at the net level. The World Economic Forum expects a net gain of 78 million jobs by 2030 (170 million created against 92 million displaced). PwC's 2025 Global AI Jobs Barometer found job numbers are still rising even in the roles most exposed to automation. The gains and losses are unevenly distributed, however, so specific occupations and workers can still be displaced even as totals grow.
Which jobs are most at risk from AI?
Goldman Sachs found office and administrative support (46% of tasks exposed), legal (44%), and architecture and engineering (37%) among the most exposed, versus construction (6%) and maintenance (4%). The World Economic Forum lists cashiers, administrative assistants, bank tellers, and data-entry clerks among the fastest-declining roles. Pew Research notes that many white-collar analytical jobs, such as budget analysts and tax preparers, are also highly exposed.
What percentage of work could be automated by AI?
McKinsey Global Institute estimates that up to 30% of the hours worked in the US economy could be automated by 2030 as generative AI is adopted, up from 21.5% without it. Goldman Sachs estimated that AI could substitute for roughly a quarter of current work tasks, and the OECD found 27% of jobs are in occupations at high risk of automation.
How many workers will need reskilling because of AI?
The World Economic Forum's Future of Jobs Report 2025 estimates that 59% of the global workforce will need reskilling or upskilling by 2030, that nearly 40% of the skills required on the job will change, and that 63% of employers see the skills gap as their biggest barrier. Some 85% of employers plan to prioritise upskilling their workforce.
Does AI increase or decrease wages?
Evidence so far points both ways. PwC's 2025 Global AI Jobs Barometer found that jobs requiring AI skills carry an average 56% wage premium and that wages are growing about twice as fast in AI-exposed industries. At the same time, the International Monetary Fund warned in 2024 that AI is likely to worsen overall inequality, because higher earners and cognitive-task roles are the most exposed to both its benefits and its risks.
Planning operations around AI?
If you are weighing where AI fits against skilled human operators, Actigy runs both — routine volume handled by AI, exceptions and quality owned by trained people, against documented SLAs.