Artificial intelligence is already interfering with labor, and recent graduates may be “lost generation”
“Artificial intelligence will replace half of all white-collar workers in the United States,” a direct offer from Ford CEO Jim Farley. He wasn’t the only executive who sounded the alarm.
Confused CEO Aravind Srinivas recently told The Verge that he hopes AI can Replace recruiters and executive assistants In the next six months. The warning is enough, and the timeline given by executives is relatively short. But reality may be more imminent.
“The disruption of work is already underway, it is expanding rapidly and will continue to grow,” said John McCarthy, associate professor of global labor and work at Cornell University’s School of Industry and Labor Relations.
The latest AI assistant for anthropomorphism, Release on July 15, doing almost everything a financial intern will do at an average Wall Street company. Tobias Lütke, CEO of Shopify Internal memo Earlier this year. Luis von Ahn, CEO of Duolingo Similar memos were sent to workers this year.
“We can certainly say that we are facing a serious collapse in the early stages of the white-collar career, which is really important because that’s the beginning of economic security and where we really build the foundation. Now, this foundation is starting to be evacuated,” McCarthy told Gitzmodo.
Young graduates may be “lost generation”
It’s a particularly bad time now that it’s an unemployed 20s.
New York Federal Reserve released Report April said the labor market for college graduates aged 22 to 27 has “deteriorated significantly in the first quarter of 2025” and the unemployment rate is the highest since the pandemic. The recent unemployment gap between graduates and all workers is being Widest Since 1990.
Some of these are related to broader market-level trends, the end of post-cultivation recruitment boom and economic softening, but AI remains a significant factor. Generative AI is particularly good at basic tasks that may be expected to be completed as an entry-level worker.
“The evidence shows that AI has a strong negative impact on early careers, and I’m worried that the current generational squeeze could evolve into a permanent reconfiguration of the early career path,” McCarthy said.
In fact, this is a tear-up of recent graduates’ social contracts: entry-level white-collar work should be the training route for the rest of your career. For recent college graduates, we may see these opportunities, and we may see (as McCarthy sees) increasing reliance on elite internships and networks. This will only expand inequality.
“People do worry that unless policy, education and hiring norms adapt, I have an entire cohort, that is, people who graduated during the early transition of AI, probably a generation,” McCarthy said. “And I’m not very optimistic about the adjustments to the scale they need.”
Is it a scapegoat?
However, Robert Seamans, professor of management and organization at NYU, believes that we are not in the circumstance of the labor crisis because despite the hype and hiring freezes, we actually see “relatively low AI adoption rates” across the company’s department.
According to the recent Feeding paperone of the biggest challenges in scaling AI now is not the technology itself. It enables businesses to actually use it. The paper says that most companies outside the technology, finance and science industries have not yet played the generative AI of AI in their daily operations, and even in larger companies, adoption rates are much higher than smaller companies.
“Implementing AI in a company is much more difficult than people realize,” Seamans told Gizmodo. “Companies usually don’t have internal talent to train, operate and monitor any AI they implement, so it’s hard to rely heavily on AI until you have someone with expertise.”
Instead, seafarers believe that some of these companies that are freezing recruiting or shrinking employee AI are actually scapegoats for using technology as a scapegoat for company performance.
“It’s much harder to blame tariffs or economic uncertainty because there aren’t so many hiring reasons,” Seamans said.
Still, even the Seaman says he can’t help but notice that signs pointing to AI are at least part of the culprits of what’s going on in the labor market. But to better understand the role of AI, we need the opinions of “a well-funded U.S. statistical agency”, such as the U.S. Census Bureau or the Bureau of Labor and Statistics.
“I think that really emphasizes the need for the federal government to track AI deployments in real time, so we can actually do some research with the current data instead of guessing if AI is likely to be responsible,” Seamans said.
Where do we go from here?
AI stays here. Moreover, AI innovation and diffusion in the corporate world seem to be able to accelerate from here.
This doesn’t necessarily lead to a “wholesale termination” of the job, but a reorganization, McCarthy said.
“Human work is changing, it will continue to evolve, and it’s hard to predict how it is shaped, but I think there is a lasting need for roles that need judgment, morality, creativity and work that need to be combined with context,” McCarthy said.
This reorganization will put pressure on universities and even K12 institutions to prepare students accordingly, not just in computer science classes. McCarthy said he has implemented it to Cornell by teaching his students to assist workflows and tools, as well as other general skills in other classes.
The other half of the coin is policy.
“These changes are happening very quickly and have greater potential than any little bit of impact on work in history, and I think there is a pressing need for multifaceted dialogue on every level,” McCarthy said. He added that public policy, educational institutions and the private sector should continue to talk to each other in terms of addressing these issues.
For workers who want to adapt, McCarthy says the most important thing is to use AI tools to improve your adaptability to your new role and be ready to pivot.
“I won’t say these things easily. Unfortunately, I don’t think any of these changes are easy or comfortable,” McCarthy said. “I have a 7-year-old and I’m very worried about what his job will look like for him in the future.”
