Artificial intelligence (AI) and AI-powered robots are now capable of performing more than half of all work hours in the United States, according to a new report from the McKinsey Global Institute.
The report found that although the vast majority of human skills will remain relevant in an era of large-scale automation, the way people use those skills is expected to change dramatically.
McKinsey’s analysis, published Nov. 25, concludes that today’s AI systems and workplace robots could automate 57 percent of U.S. work hours without any further breakthroughs in technology, provided companies redesign their workflows around automation. At current levels of capability, AI agents could perform tasks that occupy 44 percent of U.S. work hours today, and robots could account for 13 percent, the report said.
“Extending automation further would require technologies that can match a range of human capabilities currently unmatched,” McKinsey analysts said in the report. “Agents would need to interpret intention and emotion. Robots would need to master fine motor control, such as grasping delicate objects or manipulating instruments in surgery.”
The report shows that nearly 40 percent of U.S. jobs are occupations that involve daily tasks that could be automated by software alone, such as administrative support, paralegal work, office roles, and certain programming jobs.
“Tasks occupying more than half of current work hours could potentially be automated, primarily by agents,” the analysts state.
U.S. government projections show flat or declining employment in several tech-exposed fields—including administrative support occupations, paralegals and legal assistants, and computer programmers—and the Bureau of Labor Statistics (BLS) has cited technology, including AI, as a major factor.
“Software innovations have automated many of the tasks performed by bookkeeping, accounting, and auditing clerks,” the BLS said in a note. “As a result, the same amount of work can be done with fewer employees, which is expected to lead to job losses over the projections decade.”
AI Reshaping Jobs
McKinsey analysts said that a human worker won’t necessarily be replaced by AI in a given job just because the machine can do it.
“Yet, that does not mean half of all jobs would disappear; many would change as specific tasks are automated, shifting what people do rather than eliminating the work itself,” they wrote.
Instead of replacing workers outright, AI is more likely to reshape jobs by absorbing routine or time-consuming tasks while leaving humans to perform higher-value work, they said. Workers would spend less time gathering information or preparing documents and more time guiding AI tools, reviewing outputs, making decisions, and handling tasks that require judgment, empathy, or improvisation.
“A skills partnership is emerging: Machines handle routine tasks while people frame problems, provide guidance to AI agents and robots, interpret results, and make decisions,” the analysts stated. “The work blends collaboration and oversight, as humans bring judgment and contextual understanding that machines still lack.”
McKinsey identifies a set of “people-centric” roles—such as nursing, caregiving, building repair, and front-line maintenance—that remain largely resistant to automation because they rely on physical presence, dexterity, and emotional intelligence.
An accompanying technical appendix estimates that about 70 percent of caregiving tasks still require hands-on human abilities that current AI agents and robots cannot replicate. Similarly, maintenance, construction, and other jobs carried out in unpredictable physical environments show low technical automatability; more than 80 percent of their hours involve physical tasks robots cannot yet replicate.
The report also found that AI could create significant economic gains if companies redesign entire workflows rather than automating individual tasks. McKinsey estimates that large-scale adoption could generate up to $2.9 trillion a year in additional U.S. economic value by 2030, driven by productivity gains as humans and machines work together.
Other research suggests bumpy AI adoption and questionable benefits.
A Sept. 22 analysis by the Harvard Business Review found that many companies are seeing early declines in productivity as employees rely excessively on generative AI to produce “low-effort” drafts that require colleagues to redo or reinterpret the work, a pattern researchers labeled “workslop.”
A Penn Wharton report published on Sept. 8 forecast that AI could raise long-term productivity, boosting U.S. gross domestic product by 1.5 percent by 2035 and 3.7 percent by 2075, although the gains fade as industry structures adjust.
Naveen Athrappully contributed to this report.
