Concerns are brewing about AI causing people to lose their jobs: Reuters/Ipsos poll Starting in August 2025, 71% of Americans are worried that artificial intelligence will lead to permanent unemployment. last week, Amazon Announce The company will cut 16,000 jobsa total of more than 30,000 layoffs since October 2025. The move coincides with Amazon’s push to develop artificial intelligence, although the tech giant attributes the layoffs to an attempt to cut bureaucracy rather than technology.
the latest one Report Research from the Yale Budget Lab suggests there is truth to Amazon’s assertion that these mass layoffs, even at tech companies, are not the result of artificial intelligence replacing workers.
“While there are widespread concerns about the impact of AI on today’s labor market, our data suggest that this remains largely speculative,” the report said. “Our data suggest that AI’s impact on the labor market largely reflects stabilization rather than significant disruption at the economy-wide level.”
To measure the impact of AI on the workforce, the Yale Budget Lab tracked changes in Americans’ occupational mix, or job types, and the length of unemployment in jobs with high exposure to AI.
The report said that while the career mix has changed since ChatGPT was released in 2022, the rate of change is not fast enough to indicate a dramatic shift. Furthermore, the length of unemployment for individuals in jobs with high exposure to AI remains constant over time. Both indicators show no evidence that artificial intelligence or other factors are causing large-scale workforce disruption.
“No matter which way you look at the data, at this moment, there doesn’t appear to be a significant macroeconomic impact,” said Martha Gimbel, executive director and co-founder of the Yale Budget Lab. wealth.
“AI washing” in action
The Yale Budget Lab’s assertion comes against the backdrop of other data that some have interpreted as a harbinger of large-scale workforce disruptions. one with report A report released in November 2025 found that current artificial intelligence systems can already complete the tasks of nearly 12% of the workforce. Goldman Sachs Forecast Represents 6% to 7% of the U.S. workforce If AI technology becomes widely adopted, it may be replaced.
Despite growing concerns about AI-related job losses, these projections do not reflect today’s situation. The discrepancy between anxiety over job displacement by AI and data suggesting otherwise raises questions about “Artificial intelligence washing” or falsely attributing AI to company layoffs.
one Oxford Economics report Data cited last month from employment consulting firm Challenger, Gray & Christmas supports this view: While 55,000 U.S. layoffs were attributed to AI in the first 11 months of 2025, that accounted for only 4.5% of the total reported layoffs. By comparison, job losses due to standard “market and economic conditions” totaled 245,000.
“We suspect that some companies are trying to dress up layoffs as good news rather than bad news, such as past over-hiring,” the Oxford report said.
Kimbell of the Yale Budget Lab said one of the reasons companies attribute layoffs to AI is to avoid telling investors that the company is having trouble dealing with reduced immigration, tariffs and other policy uncertainties that will inevitably shake up the workforce. Anxiety associated with artificial intelligence has made the technology a convenient scapegoat for CEOs when facing skeptical investors.
“If you were a CEO, what would you say? ‘Hey, I’ve been a really bad CEO. I’ve totally mismanaged the macroeconomic situation over the past few years, so now some of you are going to have to lose your jobs, but shareholders are supposed to continue to trust me going forward?'” Gimbell said. “No, you don’t say that. You say, ‘The world is changing rapidly and we are going to right-size the company and invest forward so that we can be the most productive version of ourselves and win the future.'”
What is happening in the labor market?
She noted that it is more realistic to attribute the low hiring, low firing labor market conditions to the myriad political factors shaking the economy and the hangovers of the economic crisis. Pandemic-era hiring surge and Fed rate hike cycle This naturally slows down the development of the job market.
To be sure, economic constraints may affect how quickly new technologies are implemented, Kimbell said, providing a blueprint for when AI will begin to have a greater impact on the workforce. For example, during the first industrial revolution, Trade restrictions of the Napoleonic Wars This resulted in factory owners rushing to invest in technologies such as power looms and spinning jenny, which could automate weaving and replace workers.
“Technological change doesn’t happen in a vacuum,” she said.
Kimbell said the next big test for AI in the labor market will be if a recession hits, requiring changes to incentivize mass adoption of AI. AI adoption and productivity gains, according to PwC always very humble56% of companies said they have yet to “get any benefit” from AI.
Kimbell said that if AI brings significant changes to the job market, this will be reflected in large changes in the mix of jobs people hold and in the length of time they are unemployed for those who had more exposure to AI in their previous jobs. Otherwise, she noted, now is not the time to sound the alarm.
“If you think the AI apocalypse is coming for the labor market, it’s not helpful to announce that it’s already here before it’s here,” she said. “That could all change. That’s why we’re tracking it… Just because a technology can do something doesn’t mean everyone’s going to lose their job tomorrow. It doesn’t mean they won’t lose their job in five years, though.”