Automation will erase labor demand equivalent to 36 million U.S. jobs by 2035, while other engines of the economy generate roughly 41 million, according to a new McKinsey Global Institute report. The net balance is positive. The distribution is not: around 11 million workers — about 7% of the current workforce — will need to leave their occupation entirely, and nearly half have no clear route into a new one.
- A net gain, unevenly shared: The 36 million roles affected represent 21% of current work hours, with a projected net addition of roughly 5 million jobs.
- Transitions must accelerate 3.6x: Some 770,000 workers a year would need to switch occupational groups, against a historical average of 215,000.
- Credentials are the bottleneck: Only 14% of displaced workers have a direct path to a growing occupation; 45% face what the report calls an “unpaved” route.
- Low-wage workers carry the burden: They are 7.6 times more likely to need a new profession than high earners.
What the McKinsey Report Actually Says
The 82-page study, titled Workforce in motion and published on September 29, 2026, was written by María Jesús Ramírez, Kweilin Ellingrud, Tanguy Catlin, Diego Castresana and Anna Kortis. Its central claim is that the defining challenge of the coming decade is mobility, not scarcity.
The headline figure needs unpacking. Those 36 million jobs are not 36 million layoffs. Roughly 25 million workers would remain in their occupation with redesigned tasks as sector growth offsets automation. The remaining 11 million — a range of 6 to 16 million depending on adoption speed — must change professions.
“The good news is that the demand is going to be higher than the decline,” said Tanguy Catlin, a director at the McKinsey Global Institute.
The required pace is unprecedented: about 770,000 people a year would switch occupational groups, 3.6 times the historical average of 215,000.
The Uncomfortable Number: 45% Face an Unpaved Path
Only 14% of those who must switch — one in seven — have a direct pathway into an expanding field. Another 41% face a winding route, and 45% confront an unpaved one: relocation, substantial reskilling or fresh credentials, often for lower pay.
The burden is not evenly spread. More than 70% of disappearing employment sits in the two lowest wage quintiles, while about 60% of the growth lands in the top two. Roughly 85% of growing occupations require a credential, and around 76% cannot be performed remotely — they are in hospitals, construction sites and data centers. One consolation: only about 3% of transitions, some 305,000 people, would involve a pay cut.
“To get those credentials may require an investment of time and an investment of money, and that will be source of friction,” Catlin said.
Automation Versus Augmentation: Which Jobs Change
Office and administrative support is the most exposed category, with about 81% of work hours affected and a 38% reduction in labor demand across 18 million full-time employees. Agriculture, fishing and forestry follows at 76% of hours, technology and analytics at 70%, manufacturing and production at 68%, and retail and sales at roughly 68%.
Public safety sits at the other end, with 28% of hours affected, and healthcare practitioners at 26%. More than 75% of roles needing transition cluster in three families: administration, sales, and transportation and logistics.
Exposure, however, is not replacement. Automation may touch 47% of work hours in the legal field, yet projected demand loss is only 18%, because lawyers spend less time on research and more on tasks requiring human judgment.
“Technology determines what is possible, but managerial choices determine how quickly those possibilities become reality,” the report notes.
Software illustrates the mechanism: developer postings have climbed roughly 15% since early 2025 after falling from a 2022 peak, with demand strongest for senior, AI-oriented profiles.
Where the Safe Ground Lies
A separate analysis by Hamilton Group, which ranked 60 professions carrying an automation risk below 45%, using Anthropic Economic Index data, found that nine of the ten least vulnerable occupations sit in healthcare or social services. Emergency medical technicians and paramedics scored 0% replacement risk; physicians 4.5%, with a median wage of $266,000 across 343,000 workers; mental health and substance abuse social workers 4.5%; registered nurses 6.9%. AI still assists heavily — the healthcare practitioner category shows a 58.4% augmentation rate — and the U.S. Bureau of Labor Statistics projects 6% employment growth for EMTs and paramedics through 2035.
Reading the Forecast With Caution
Context matters here. McKinsey sells transformation services to the same companies that will have to manage these transitions, and its framing of mobility over scarcity is notably convenient. In 2023 the firm projected 12 million occupational changes by 2030; it now says 11 million by 2035 — nearly identical in magnitude, with the deadline pushed five years out. The data also covers the United States only and cannot be transplanted onto labor markets with different sectoral structures and age pyramids.
The number that deserves attention is neither 41 million nor 36 million. It is the 14% with a direct path forward. Everything else amounts to training policy still unwritten — and no AI model will fix that.