What Jobs Will AI Replace in the Next 10 Years?
Most answers to this question hand you a single confident percentage. That's the wrong frame. Some of what's coming over the next decade is close to certain because it's already measured in current data. Some of it is a reasonable bet based on named forecasts. Some of it is closer to speculation dressed up as a projection. Sorting by confidence, not by date, is a more honest way to read this than any single number.
Near-certain: already showing up in real data
This isn't a forecast. Stanford economist Erik Brynjolfsson and colleagues analyzed ADP payroll data covering 25 million US workers and found a 13% relative decline in employment for 22-to-25-year-olds in the most AI-exposed occupations since generative AI adoption accelerated, while older workers in the same roles held steady or grew (Fortune). This is happening now, concentrated in entry-level software and customer service roles, and it's the most reliable data point on this page because it's measured, not modeled.
Layoff tracking backs this up at the announcement level. Challenger, Gray & Christmas, which has tracked AI as a distinct layoff reason since 2023, recorded AI cited in 173,568 cumulative job cut announcements through mid-2026, with AI leading all cited reasons for layoffs in June 2026 at 31% of that month's cuts (Challenger, Gray & Christmas). That's a real, running count, not a decade-out estimate.
Entry-level hiring is also already contracting for a named reason. A Resume.org survey of nearly 1,000 US business leaders found 51% said their company will lay off existing workers in 2026 specifically because AI is consolidating or eliminating roles, and a separate Resume.org survey found 1 in 5 companies have stopped hiring entry-level workers because of AI (PRNewswire).
Plausible: named forecasts with real methodology behind them
The World Economic Forum's Future of Jobs Report 2025, based on a survey of over 1,000 major employers, projects 92 million jobs displaced globally by 2030 against 170 million created, and found that by 2027, nearly half of companies expect to have eliminated entry-level hiring altogether, with 21% naming AI as the sole reason (World Economic Forum).
McKinsey Global Institute modeled a rise in automated US work hours from 21.5% to roughly 29.5% by 2030, attributable specifically to generative AI, with lower-wage workers facing up to 14 times more pressure to change occupations than the highest earners (McKinsey). Goldman Sachs modeled task-level exposure and put office and administrative work at 46% and legal work at 44%, against much lower exposure for physical and outdoor jobs (CNBC). Gartner, working from current enterprise deployment data rather than long-range modeling, forecasts that by 2027, half of companies that attributed headcount cuts to AI will rehire for similar functions, because they're hitting real limits in complex problem resolution and contextual judgment (Gartner via TechRepublic). These are credible, named, methodologically different forecasts that mostly agree on direction: routine, low-accountability, digital-only work is under real pressure across the decade.
Speculative: repeated often, verified rarely
Any claim that names a specific percentage of "all jobs" disappearing within 10 years, without naming a source and a method, falls in this bucket. The most-cited example of this failure mode is instructive: in 2013, Oxford researchers Carl Benedikt Frey and Michael Osborne estimated 47% of US jobs were at high risk of automation "perhaps in a decade or two" (Oxford Martin School). More than a decade later, US total employment had grown, not shrunk, and a later study out of the University of Mannheim, using a more granular task-based method, put the comparable US figure closer to 9%, not 47% (ITIF). Frey and Osborne's own paper never claimed a pace of automation. It measured technical exposure, and the press coverage that followed turned that into a doomsday number the study itself never made.
That pattern repeats. Jeremy Rifkin's 1995 book "The End of Work" predicted a labor market breakdown that instead coincided with 25 years of nearly continuous US employment growth, from 124 million jobs in 1995 to 158 million by early 2020 (The Conversation). ATMs were widely predicted to wipe out bank tellers. They didn't. The lesson isn't that automation forecasts are always wrong. It's that a specific date attached to a sweeping number is the least reliable part of any forecast, including the ones cited above as "plausible." Treat direction as more trustworthy than the calendar year attached to it.
What holds across all three tiers
The pattern that survives from measured data through to plausible forecast is consistent: routine, digital-only, low-accountability tasks face rising pressure, and that pressure lands hardest on entry-level and early-career workers first. Nothing in the near-certain or plausible tiers supports a claim that most jobs, or even most of any single occupation, disappear within 10 years. What the data supports is that the easiest, most repeatable slice of many jobs gets automated first, and that slice is often exactly what junior employees are hired to do.
Keep the judgment layer yours
The tasks you can still do without leaning on AI are what make you hard to replace here. The free 5-Day AI Reset is a five-email course built around exactly that: Day 2 has you take one task back and do it unassisted. One small change per day, and it stays useful no matter which way in the next 10 years moves.
Frequently asked questions
What's the most reliable data on jobs AI will replace in the next 10 years?
Real payroll data beats any model. Stanford's analysis of 25 million ADP records found a measured 13% relative decline in entry-level employment in AI-exposed fields, which is happening now rather than being projected for later (Fortune).
Is the "47% of jobs automated" statistic still accurate?
No. That 2013 Oxford estimate by Frey and Osborne was a measure of technical exposure, not a timed prediction, and a later, more granular study put the comparable US figure closer to 9% (ITIF). It's the most-cited example of a forecast losing its caveats in public retelling.
Will entry-level jobs recover after AI-driven cuts?
Gartner forecasts that by 2027, half of companies that cut headcount citing AI will rehire for similar functions, often under different titles, as they hit limits around complex problem-solving and human judgment (Gartner via TechRepublic).
Which jobs face the least pressure over the next decade?
Physically demanding, outdoor, and high-accountability work, per Goldman Sachs' task-exposure modeling (CNBC). Roles requiring licensure, legal accountability, or in-person physical presence consistently rank lowest across every forecast on this page.
Should I trust a 10-year jobs forecast at all?
Trust the direction more than the number. Every credible forecast above points the same way on routine, low-accountability work. None of them, including the credible ones, should be read as a precise headcount for a specific year.