Artificial intelligence could automate more than half of US working hours by 2035 without eliminating half the workforce, according to the McKinsey Global Institute. Its research estimates that technological and economic changes could reduce labour demand by the equivalent of 36 million jobs over the next decade. Meanwhile, economic expansion, including growth in AI-related industries and services, could generate demand for approximately 41 million jobs. Overall employment could therefore increase even as the composition of work changes substantially.
The principal challenge is equipping workers for roles whose requirements differ from their existing skills. In management and law, where judgement, coordination and accountability are important, AI is more likely to augment employees than replace them outright. Training may be needed for responsibilities such as reviewing AI-generated output, overseeing responsible use and redesigning workflows. Office and administrative support roles are particularly exposed: approximately 80% of their current working hours could be automated. That is not a forecast of equivalent job losses, although around 11 million Americans may need to move from declining occupations into different ones.
Qualifications present another obstacle. About 85% of growing employment requires a credential or certification, either legally or because employers prefer one. Companies could broaden recruitment by reassessing essential qualifications and prioritising demonstrated skills and experience. McKinsey recommends monitoring transitions into new roles, internal recruitment and worker outcomes alongside cost savings. Businesses should also assess whether AI’s time savings enable improved customer service, innovation, business growth or more demanding work.
