McKinsey: AI Could Absorb Half of U.S. Work Hours by 2035

AI and related technologies could automate more than half of all U.S. work hours over the next decade, according to new research from the McKinsey Global Institute (MGI). That figure does not mean half the workforce will disappear.

The report, "Workforce in Motion: Skills and Pathways to Future Jobs in the United States," quantifies the disruption in stark terms. Technological and economic changes could reduce labor demand by the equivalent of 36 million jobs over the next decade. At the same time, economic growth — including expansion in AI-related industries and services — could create demand for about 41 million jobs.

The net result: the U.S. could have more jobs in 2035 than it does today, even as technology rewrites how people work and what skills they need.

Where will automation hit hardest?

Office and administrative support roles face the greatest exposure. Roughly 80% of current work hours in those functions could be automated by 2035, MGI estimates. The figure does not mean 80% of those jobs will vanish — but it signals how dramatically the roles could change.

The picture differs in fields such as management and law, where judgment, coordination and accountability matter. There, AI is more likely to help people do their jobs than to replace them outright. The technology may also create new responsibilities that demand training: checking AI-generated work, ensuring its responsible use and redesigning workflows around it.

How many workers must switch occupations?

MGI estimates about 11 million Americans may need to leave declining occupations for entirely new ones. Others may stay in their fields as the work itself evolves.

The annual pace of change is the real headline. Roughly 770,000 workers a year may need to change occupations over the next decade — about 3.6 times the historical rate, according to the report's authors.

Do credentials block the way?

Yes, in many cases. About 85% of growing employment requires a credential or certification, MGI found. The breakdown:

• 38% of those jobs carry a legally required credential.
• Another 47% have credentials that employers prefer, though the law does not mandate them.

The research suggests companies could widen their talent pools by rethinking which qualifications are genuinely essential and giving greater weight to proven skills and hands-on experience.

What should executives measure?

MGI recommends that companies track outcomes, not just efficiency gains. Specifically:

• Whether employees actually move into new roles.
• Whether companies fill more positions internally.
• How workers fare after transitions — not just AI-related cost savings.

For finance leaders, the research poses a sharper test: examine how employees use the time AI frees up. Does that time go toward improving customer service, developing new ideas, growing the business or tackling more complex work? Or does it simply shrink the hours spent on existing tasks?

The report's central question is not how much work AI can take on. It is what the business does with the time it saves — and whether employers build the training pathways to get 11 million people into the jobs the next decade will create.