McKinsey Maps the AI Job Shift: 36 Million US Jobs Disrupted, but Most Moves Stay Within Occupations

The McKinsey Global Institute published a new report, 'Workforce in motion: Skills and pathways to future jobs in the United States,' examining how AI and automation will reshuffle American work over the coming decade. The institute's modeling projects that automation could disrupt about 36 million U.S. jobs by 2035, while the economy creates more than 40 million jobs, and roughly 11 million workers change occupations over the next decade.

The report's most important point is not the headline number but where the movement happens. What makes the coming shifts different from past waves, McKinsey argues, is not how many workers move but where they need to move. Historically, transitions happened within broad occupational groups that share similar skills, lowering barriers to mobility. The institute expects most workers in the 36 million disrupted jobs to follow that same pattern and transition into a job in the same occupation, which is a far less disruptive story than mass occupational switching.

The practical implication is a shift from managing job titles to managing skills. McKinsey describes companies sorting employees into those able to migrate quickly into growing roles, those needing targeted upskilling, those requiring longer transition pathways, and roles that cannot be filled internally. Recent McKinsey research finds companies expect to retrain nearly one-third of their workforce in response to AI-driven skill mismatches, complemented by external hiring and mobility.

A caveat worth keeping in mind: these are modeled projections, not observed outcomes. Automation adoption rates have repeatedly come in slower than forecasts, and the 2035 numbers depend heavily on assumptions about how fast AI capabilities translate into workplace deployment.

Why it matters: the report reframes the AI labor debate from mass displacement to managed reallocation. That changes who is responsible, employers for internal mobility and reskilling, community colleges and workforce agencies for the longer transition pathways, and workers for accumulating adjacent skills. The binding constraint is pathways, not headcount.