The job market is beginning to shake. How much responsibility does artificial intelligence take?
Not much. At least not yet.
Reviews of employment surveys, interviews with labor market analysts, and recent corporate revenue reports have shown that so far little evidence supports claims of the widespread economic impact of the growing use of AI.
“We're accusing the company of the world,” said Martha Guin Bell, executive director and co-founder of Yale University's Budget Lab and former presidential economic advisor Joe Biden. “And that's not there yet.”
There's a lot to be done in the payoffs from AI. The stock market has reached its highest highs, mainly thanks to profits from tech giants such as NVIDIA, Google Parent Alphabet, Facebook Parent Meta and Microsoft.
For precisely that reason, analysts say some companies may be encouraged to hype the possibilities of AI as a disruptive force. By the end of July, the term “AI” was cited in about two-thirds of second-quarter revenue calls made by S&P 500 companies, according to data provider FactSet. That's less than half of the first quarter.
Amid the downshift in the economy, cost pressures are rising, urging corporate leaders to hype the potential of AI as a source of savings.
“In 2023, well-known public companies can reduce and quote interest rate increases or uncertain macro conditions,” said Roger Lee, a high-tech entrepreneur who runs a website tracking layoffs in the tech industry. “Today, it's AI.”
The most extreme warning about the short-term impact of AI comes from Dario Amodei, co-founder and CEO of AI Business Humanity. In May, he told Axios that he foresaw that half of the entry-level white-collar jobs would be wiped out over the next 1-5 years, raising the unemployment rate from 10% to 20%.
So far, evidence for this scenario is mixed. According to labor market analytics firm Revelio Labs, entry-level or all other jobs have been declining since 2023, but the trend has not been linear. Revelio said that entry-level jobs exposed to AI are declining the fastest, but the advanced roles exposed to AI are actually beginning to recover.
The broad picture of white-collar occupations at the highest risk of confusion actually shows a fairly stable employment trend. Last week's official employment report shows that the roles of office and management have actually returned to pandemic-era highs, but employment in other specialties such as accounting and legal services is relatively stable.
It's a pessimistic story of technology, but it's also a more subtle story when it comes to the impact of AI. Both Amazon and Microsoft leaders show their ability to run their businesses with reduced staff thanks to AI. The high-tech layoffs tracked on Lee's website hit a three-month high in July, and were the parent of Endall, the three companies (Intel, Microsoft and Recruit Holdings), and Glassdoor's parents.
All of these companies cited artificial intelligence as a role in job cuts, Lee said. However, he pointed out that in the case of Recruit Holdings there was no details on how AI had affected the lost positions. The company simply said the technology is “changing the world.”
“Many of the roles being cut seems to be consistent with those used by AI,” Lee said. “But it's still used as a cover in other cases.”
Recruiting representatives did not respond to requests for comment.
The simple calculation behind AI is that the less a company can do more. But economists say it's difficult to calculate accurate changes in productivity in the short term, but so far, the broadest national measures have shown a slowdown in recent quarters.
A future paper from Carnegie Mellon and Stanford University researchers reveal that most of the benefits of AI comes from consumers, not businesses. If a lot of the value from today's generation of AI seems to be able to make emails and papers faster or do more quickly, you don't imagine things.
“Free products can improve consumers, but they can't see in GDP count,” authors Avinash Collis and Erik Brynjolfsson wrote in the recent Wall Street Journal Op-Ed. They calculate the equivalent of $97 billion in surplus welfare from the generated AI in 2024, compared to the $7 billion revenue recorded by the tech companies that actually created the AI products.
The economy usually has a “J-Curve” effect when transformative technologies are introduced, Collis told NBC News. These initial effects are often not captured by official numbers, but at first there are some bottlenecks that can cause confusion. For example, the iPhone has increased its total total of photos from billions to trillions. This directly impacted workers at camera giant Kodak, but created immeasurable opportunities elsewhere, Colis said.
“It's probably going to have a lot of negative impacts on some sectors,” Collis said. “But at the same time, I was able to create many new jobs.”
Other indicators suggest agitation of more pronounced AI effects on jobs. A July employment survey by Consultancy Challenger, Gray and Christmas found that companies denounced 20,000 job cuts in 2025 as “automation and AI implementation,” and that an additional 10,000 people were directly attributed to artificial intelligence. The Challenger said this represents “a significant acceleration in AI-related restructuring.”
These figures are warned by reduced government spending and cuts related to general economic and market conditions, and will play a role of nearly half a million people this year, Challenger said.
Some companies appear to be investigating the possibility of addressing the widespread uncertainty of the economy to stabilize payroll and using additional resources to increase AI's bottom line profits. MoviePass CEO Stacy Spikes told NBC News that his company's internal workflow will be much more efficient thanks to AI. It has made him more guns about taking new workers to certain departments like software. As of Tuesday, no open positions were displayed on the MoviePass career page.
“We haven't seen the need to increase our staff,” Spikes said.
However, businesses like MoviePass still seem to be exceptions. Analysts at Goldman Sachs say only about 9% of all companies use new AI tools to produce their products and services. As a result, only limited effects are seen at this time.
“When I look at the impact AI has had on data across the entire labor market so far, that seems pretty small to me,” said Joseph Briggs, head of the Global Economics Team at Goldman Sachs Research, in a recent company podcast. Even recent university graduates who saw the unemployment rates said, “The relationship anecdotes and anecdotes have with AI are often a bit exaggerated,” Briggs said.
Analysts at JP Morgan have come to a similar conclusion, finding for now that the study “has not been able to significantly affect job growth.”
But they warned that this could change in the next recession.
For white-collar workers, “I think the speed and breadth of adoption of AI tools and applications in the workplace can induce a massive displacement of the occupation in the course of the next recession,” they said in a recent note to clients.
Others remain more optimistic about the possibility of new opportunities to overcome the negative effects. That's how Jensen Huang, co-founder and CEO of Nvidia, sees it. As the head of the AI giant, he may also be hyped about its potential, but his outlook is particularly rosy than humanity's Amody. Huang told Axios last month that the technology will ultimately lead to more work, even if there are some redundancy elsewhere.
“Everyone's work will change,” he said. “Some jobs are not needed. Some people will lose their jobs. But there will be many new jobs. …The world will become more productive.”
