AI might displace 11m US employees, McKinsey says

An artificial intelligence (AI) job apocalypse might not be the biggest threat to US workers in the next decade. The larger challenge could be helping millions of people move into new jobs, according to a study by consulting firm McKinsey & Co.

About 11 million US workers, about 7 percent of the workforce, might need to change jobs by 2035 because of AI-related displacement, said the report released yesterday.

That means the number of US workers who change occupations each year — historically about 215,000 — might triple over the next decade. The transition would be difficult for many, with six out of seven workers facing substantial retraining and loss of income, McKinsey said.

An AI logo is displayed during the opening day of the IFA in Berlin on Sept. 4.

Photo: EPA

Workers are not “interchangeable units,” the report says. “Job opportunities can be abundant and yet leave millions of workers without work if those positions require different skills, credentials, locations or pay structures than current jobs.”

The McKinsey study found that full-time employees most likely to have to change jobs are concentrated in three groups: Office and administrative support, retail and sales, and transportation and logistics. That includes jobs such as customer-service representatives, cashiers and warehouse workers. Many are low-income workers, whom the report says would be almost eight times as likely as higher earners to have to change jobs.

Like many projections of its kind, the analysis hinges on factors that are difficult to predict, including how quickly companies adopt AI and how that affects demand for workers across industries. The research accounts for broader trends shaping labor demand, including the country’s aging population. Given that uncertainty, the authors estimate that 6 million to 16 million workers might need to switch occupations.

Earlier labor market transformations, such as the US move away from agriculture, unfolded far more gradually, but there is precedent for this pace of change: The rate of projected job changes is similar to what US workers experienced during the COVID-19 pandemic, said Tanguy Catlin, a senior partner at the McKinsey Global Institute and one of the report’s authors.

“We did it during COVID for two years,” he said. “You would need to presume that we can sustain it for about 10 years.”

One factor the report did not consider, Catlin said, is whether demand for workers could fall faster in some occupations than it rises in others.

“We don’t know whether it’s going to be painless,” he said.

The report outlines several ways federal, state and local governments could help workers make those transitions, including providing more detailed and timely labor-market data and funding retraining programs.

“The education system has been developed over time really to focus mostly on youth,” Catlin said. “But what can be done to continue to upskill the existing workforce, people who need to change occupations? Those are the questions we wanted to tee up for public debate.”