McKinsey Says AI Will Net Out to More US Jobs by 2035, But 11 Million Workers Must Switch Careers to Get There

The AI jobs panic has a counterargument, and it comes with a big asterisk.

The McKinsey Global Institute released a report on Tuesday, September 29, 2026, projecting that AI and automation will cut demand for roughly 36 million US jobs by 2035. That's nearly 21% of existing work hours, according to Crypto Briefing's review of the report. But McKinsey also projects growth elsewhere, mostly from AI-adjacent industries and broader economic expansion, will generate demand for 40 to 41 million new jobs over the same stretch.

Net positive. Not a wipeout.

The 11 Million Workers Needing Career Switches

Of the 36 million workers facing reduced demand, McKinsey estimates 25 million can stay in roughly the same occupation because growth in their industry offsets the hit. The remaining 11 million, about 6.5% to 7% of the US workforce, will likely need to leave their field entirely, according to the report. McKinsey's range runs from 6 million to 16 million depending on how fast automation spreads. That gap matters enormously for public budgets and private training programs.

"The next decade's challenge is mobility, not scarcity," the McKinsey researchers wrote, as quoted by both Fortune and CNN.

That mobility requirement is steep. McKinsey says the US would need about 770,000 occupational transitions a year to absorb the shift, roughly 3.6 times the historical average of around 215,000, according to Bloomberg and Quartz. For comparison, about 788,000 workers a year made similar jumps during the pandemic years of 2019 to 2022 without lasting economic damage, Fortune reported.

Who Gets Hit Hardest

This isn't spread evenly. Lower-wage workers are 7.6 times more likely than higher earners to need a new occupation, and workers without a bachelor's degree are 1.8 times as likely as degree-holders to face full career switches, according to Quartz's summary of the report. The jobs shrinking fastest are concentrated in office and administrative support, retail, and transportation. Growth is concentrated in healthcare, construction, and professional and technical services.

Only one in seven displaced workers has what McKinsey calls a "direct pathway" into a growing job requiring minimal retraining and no pay cut. Nearly half face an "unpaved" path blocked by skill gaps or credentials. And 85% of growing jobs require some kind of credential or certification. This is a real bottleneck for workers who need to move fast.

It's not that jobs vanish into nothing. Rather, 11 to 16 million Americans, disproportionately lower-wage workers without degrees, may be asked to retrain into fields gated by licensing requirements and certifications they don't currently hold. A retail clerk doesn't become a nurse overnight, credential or no credential. That's a legitimate structural problem.

New York Is Already Living This

A report released Thursday, October 1, 2026, by the Partnership for New York City, a coalition of more than 300 corporate and business leaders, found AI companies leased more than 2.2 million square feet of Manhattan office space in the first half of 2026, more than double their total for all of 2025, according to the Epoch Times. New York AI startups raised a record $16.7 billion in venture capital in 2025.

But entry-level postings have cratered. Since ChatGPT launched in 2022, entry-level postings are down 40.6% in design, media, and writing; 34.4% in customer support; 30.5% in clerical and administrative work, per the Partnership's data as reported by the Epoch Times. Meanwhile entry-level postings seeking AI skills specifically rose 55%. About 49% of career paths available to workers without a bachelor's degree are already exposed to AI, the report found.

The Macro Data Tells a Calmer Story, For Now

Fox News Digital highlighted a Bureau of Economic Analysis paper and a Budget Lab at Yale study finding no significant national link between AI adoption and employment declines so far, with AI use instead tied to stronger economic growth. Fox interviewed ordinary Americans who weren't buying it. "Everybody's probably going to get fired, just wait," one Virginia resident, Ivory, told Fox News Digital.

These two pictures aren't actually contradictory. The Yale and BEA data are national and current. The Epoch Times' New York findings are specific to entry-level white-collar postings in one city's hottest AI market. Both can be true: the broad labor market hasn't cratered yet, while the first rung of certain career ladders is already getting harder to reach.

President Trump has pushed back against tech-industry calls to slow AI development, while Senator Bernie Sanders has called for a pause, and former President Barack Obama has echoed concerns about the pace of change, according to Fox News Digital.

The Backdrop Is Already Rough

Separate from AI, the labor market is stuck. Bureau of Labor Statistics data released Tuesday showed job openings at a five-month low as of the end of August, voluntary quits near a six-year low, and layoffs shrinking for a second straight month, CNN reported. Glassdoor's Employee Confidence Index hit a record low in September, its third such low this year, according to Glassdoor chief economist Daniel Zhao. The Conference Board's consumer confidence index dropped 6.7 points to 81.9 in September, a 12-year low, per Quartz.

McKinsey didn't design a government retraining program, and nobody's funding one at the scale the report implies is needed. Whether 770,000 Americans a year actually get the credentials, the retraining, and the employer willingness to make that jump is the open question McKinsey's own numbers don't answer.