Artificial intelligence is beginning to shape the labor market in ways that worry leading economists. JPMorgan has warned that white-collar knowledge workers could be at the center of a major shift.
Murat Tasci, senior US economist at JPMorgan, believes AI could replace a wide range of office staff. These roles include “non-routine cognitive occupations,” which cover much of America’s white-collar workforce.
That risk could lead to what experts call a “jobless recovery.” In such a case, job growth remains weak, even when the economy shows signs of rebound, leaving many educated workers sidelined.
Tasci said white-collar workers may face a structurally higher risk of joblessness in the years ahead. He stressed that their slow recovery from downturns could create lasting strain in the economy.
The problem is important because knowledge workers account for nearly 45 percent of household employment in the United States. Any long slump in their job market could hurt growth across sectors.
“A much larger unemployment risk and anemic recovery prospects for these workers might cause the next labor market downturn to look pretty dismal,” Tasci wrote in a recent client note.
The bank added that policymakers may need to step in if trends worsen. Stimulus packages or softer monetary policy could be required to help workers adjust to rapid change in their industries.
Concerns about AI displacing white-collar jobs have been present on Wall Street for several years. But economists say recent labor data may show the first real signals of this pressure.
Entry-level roles appear to be the most at risk. Recent college graduates are facing higher rates of joblessness, which JPMorgan has linked to the surge in AI adoption across business sectors.
The bank also pointed out that unemployment among non-routine cognitive workers now exceeds that of routine workers. This marks a sharp break from past patterns of job market downturns.
Charts provided by JPMorgan show how the share of unemployment for non-routine cognitive jobs has steadily risen. Routine workers now make up a smaller portion of the unemployed population.
Routine occupations, both cognitive and manual, have already faced similar disruptions. Workers in sales or repetitive manual roles lived through earlier “jobless recoveries,” where many jobs never returned after recessions.
Over the past four decades, routine work in the US dropped from about 55 percent of total jobs to 40 percent. The decline reflects both automation and shifts in the broader labor structure.
So far, the impact of AI on the national job market has not pushed up headline unemployment. The overall jobless rate in the US stood near 4.2 percent in July, close to historic lows.
But JPMorgan economists say pressure could build gradually. White-collar workers may see job security fall over time, even if national numbers appear stable in the short run.
Tasci noted that routine jobs historically show a strong cyclical pattern. Each recession over the past 40 years made it harder for those roles to bounce back after the downturn ended.
“We are not seeing an imminent downturn in the labor market, though the risks are higher relative to a month ago,” Tasci said, urging close watch on upcoming economic data.
Recent hiring numbers may reflect that early weakness. The US added far fewer jobs in July than analysts expected. Job gains from May and June were also revised downward by 258,000 combined.
This weaker trend adds weight to concerns about the fragility of the labor market. If white-collar unemployment continues to climb, growth may slow even as the economy avoids a technical recession.
Economists say the United States has already lived through major transitions in work caused by automation. But the scale and speed of AI’s effect on office jobs could prove more disruptive.
The challenge ahead will be how policymakers respond if AI reduces demand for educated labor. The wrong moves could leave millions without clear paths back into stable, long-term employment.
For now, the economy remains steady, but risks are rising. As Tasci noted, the coming years may define whether AI delivers shared prosperity, or whether it drives a long period of uneven recovery.
