The rise of artificial intelligence may cut demand for millions of American jobs, a new forecast suggests it will coincide with even more job creation — provided workers can move between them.

According to a report from the McKinsey Global Institute released on September 29, 2026, AI and automation could reduce demand for about 36 million US jobs by 2035, while growth in other parts of the economy creates roughly 41 million new positions — a net difference of roughly five million. The difficulty, the research arm of McKinsey & Company finds, is that the people losing work and the jobs being created may not line up neatly.

The math works. The logistics are another story.

The 36 million jobs facing lower demand account for nearly 21% of existing work hours in the United States. On the other side of the ledger, McKinsey projects 40 to 41 million new positions, attributing most of that growth to broader economic expansion and activity across the AI value chain.

Not everyone affected will need to start over. Of the 36 million jobs facing reduced demand, about 25 million workers are likely to stay in related occupations, though with different day-to-day duties. The remaining roughly 11 million workers — around 6.5 to 7% of the labor force — may need to move into entirely new occupations, because adjusting their current roles will not be enough. McKinsey's estimate for that group carries a wide range: somewhere between 6 million and 16 million workers.

A pace the US labor market has never sustained

To absorb the required career changes, McKinsey estimates the US would need approximately 770,000 occupational transitions per year. In this context, a transition means changing fields, not just switching employers — and the 2035 horizon leaves roughly a decade to make it happen. The historical average is 215,000, meaning workforce mobility would have to run at about 3.6 times its usual annual rate.

The firm frames the challenge as one of skills alignment rather than a shortage of jobs. The report describes the shift as potentially the largest and most sustained workforce transformation in US history.

Who is exposed and who is hiring

McKinsey finds that lower-wage workers and people without college degrees are the most vulnerable to displacement. The sectors flagged as most affected include office administration, retail, sales, and transportation — categories that rank among the most common jobs in the US economy, which is part of why the exposure figures reach into the tens of millions.

On the growth side, the report expects the strongest job gains in healthcare, professional and technical services, and construction. The projections also factor in demographic pressure, with McKinsey pointing to an aging population and lower immigration rates as forces shaping labor supply alongside technological change.

What this means for employers, investors, and policymakers

The range in McKinsey's own estimate is worth keeping in mind: a world where 6 million people need new careers looks very different from one where 16 million do, both for public budgets and for private training businesses.

The metric to watch is the annual rate of occupational switching. If it starts climbing from the historical 215,000 toward the 770,000 McKinsey says is needed, the transition is working. If it stays flat while automation spreads, the net job gain on paper may not feel like a gain for the workers left behind.

Source: CryptoBriefing