No one can agree on whether AI is taking away jobs.

Recent studies have shown AI having a positive impact on job growth and opportunities, yet large groups of economists, as well as labor activists, warn that the emerging technology threatens to quickly transform the financial system, and that action must be taken now.

Tech companies, especially large ones, have continued to cull jobs during the AI boom. Microsoft laid off nearly 5,000 people in early July as it continues to pour billions into AI data centers. The layoffs added to earlier downsizing by the software giant and moves by companies that include Amazon and Oracle to shed thousands of people in the last two years.

But whether AI is directly leading to job cuts has been difficult to measure, and the picture is blurred by corporate whiplash: CEOs blame AI for layoffs one month, then hail it as an engine for new job creation the next. Even recently, some of the largest companies, such as Google parent Alphabet, have reportedly told investors they plan to increase headcount.

“There’s been discretion out there as to what extent the layoffs we have been observing are really driven by AI,” Till Von Wachter, a professor of economics at the University of California, Los Angeles, told Fortune. “It’s been notoriously hard to pin that down.”

The latest U.S. jobs report, which revealed that employers unexpectedly cut 23,000 jobs in July, has only added to the confusion.

Some economists, such as Ben Zipperer from the Economic Policy Institute, said AI’s impact on jobs has so far been more limited than what some doomsday scenarios initially predicted.

The latest U.S. jobs report, which revealed that employers unexpectedly cut 23,000 jobs in July, as only added to the confusion.

And some recent data has shown a bullish picture. A recent study by financial services firm Ramp of more than 21,000 U.S. firms found that companies that invested in AI grew their headcount. Ramp categorized its heaviest AI spenders as “high-intensity” adopters. Over two years, these top spenders expanded their overall staff by 10% and boosted entry-level hiring by 12%, defying other reports that college graduates face a barren job market. By contrast, the bottom two-thirds of adopters saw no headcount growth at all.

Though the study found general AI adopters tended to be larger firms, the most intense adopters were smaller companies which might already be growing regardless of AI and are more open to experimentation. High-intensity companies were utilizing more advanced tools like coding agents or APIs (protocols that allow various applications to communicate).

The Ramp study has parallels to a recent report by researchers at Google that found AI so far is mostly being used as a collaborative tool rather than an outright job-replacer.

Meanwhile, a June California Policy Lab study found no statewide spike in unemployment insurance claims among AI-exposed roles like software developers and customer service reps since ChatGPT’s release in late 2022, but it did find elevated UI claims specifically for college-educated workers in highly-exposed roles, as well as a significant increase in claims from high-exposed roles in the San Francisco area.

The Big Tech companies “definitely overhired during the pandemic and are now making the decisions to correct that overhiring,” Ara Kharazian, lead economist at Ramp, told Fortune. Some are “blaming it on AI. But what we’re seeing from firms that are using AI that didn’t have that overhiring problem is that they’re continuing to grow.”

He added that although many firms in his study were fast-growing to begin with, they grew even faster following AI adoption.

AI washing or AI cloaking?

Untangling AI’s true impact could take years, stymied by a phenomenon researchers call “AI washing,” where companies attribute layoffs to AI to seem forward-thinking, or the opposite trend, where companies avoid mentioning AI for fear of public outcry. Much existing research has had to rely on estimating which tasks could potentially be accomplished by AI, or indirect surveys rather than actual spending or usage records.

Some companies may also be hiring at the same time others are displacing workers, according to UCLA’s Wachter.

Yet despite some positive results, both economists and on-the-ground workers are calling attention to the negative impact AI is already having on the labor market. In July, nearly 200 economists and researchers published a statement warning that AI could cause large-scale job displacement in the next decade.

“This could drive an unprecedented transformation of our economy, larger than the Industrial Revolution, but unfolding over a vastly shorter time frame,” the statement says. Its signatories include Anthropic co-founder Jack Clark and Eric Schmidt, Google’s former chief executive. The statement calls on policymakers to “act now” to better understand how AI is transforming the economy and to create legislation that will “steer A.I. in a direction that complements humans and benefits society.”

A 2025 report co-authored by Stanford economist Erik Brynjolfsson, one of the organizers of the recent statement from economists and researchers, analyzed ADP workforce data and found that workers aged 22–25 in AI-exposed roles such as software engineering suffered a 16% relative employment drop compared to less-exposed peers.

Responding to the Ramp study, Brynjolfsson in June wrote on X that firms that adopt AI “may grow by gaining market share from non-adopters, so employment can rise among adopters even as exposed occupations shrink economy-wide.”

Kharazian, the Ramp economist, said when looking outside the high-intensity, high-growth part of Ramp’s study, the company didn’t find job gains, but it also didn’t find broad job loss.

Rank-and-file employees remain worried about the impact of AI, a spokesperson for Amazon Employees for Climate Justice, an advocacy group of current and former Amazon employees, told Fortune. Amazon Chief Executive Andy Jassy said about a year ago that AI would lead to a leaner workforce, but in February said that AI could ultimately fuel job creation, and he has framed Amazon’s layoffs as an attempt to flatten its organizational structure. The company cut about 30,000 jobs between the end of 2025 and the start of this year.

Far from making work easier, employees are feeling a “huge increased pressure” from Amazon executives to finish tasks faster using AI, the ACJ spokesperson said. AI tools have also made the demand for output higher.

An Amazon spokesperson said the company expects employees “to use all available resources—including AI tools—to help them be even more effective and have an even bigger positive impact on our customers’ lives” but said AI has not been the reason behind the majority of its layoffs, that AI adoption isn’t a factor in deciding layoffs, and that while some roles may be reduced, entirely new categories of jobs will emerge.

Amazon, Microsoft, and Oracle have all laid off thousands of workers as they spend billions on AI infrastructure. In a June filing, Oracle said its layoffs of thousands during the past year were tied to AI.

Changing their tunes

In addition to Amazon’s Jassy altering his messaging on AI killing jobs, other high-profile CEOs have softened previous comments. OpenAI Chief Executive Sam Altman had long predicted that AI would lead to huge changes in the workforce, but in May said the company had been wrong about how much “people would continue to be at the center of everything.”

Anthropic Chief Executive Dario Amodei in June wrote that his earlier comments about AI eliminating jobs were not meant to be a “prophet of doom,” but rather a call for policymakers to plan and adapt. Mustafa Suleyman, the head of Microsoft’s AI lab, earlier this year predicted that most tasks that involve “sitting down at a computer” would be fully automated by AI within the next year or 18 months, though he later tempered his stance.

Industries may handle their workforces differently based on a variety of factors. Some companies have taken aim at middle managers as the companies seek to be more nimble.

Dave Clark, founder of AI logistics startup Auger and a former senior Amazon executive, said with the help of AI, his team of roughly 80 engineers now performs with the velocity of 800 engineers.

While some data highlights risks for early-career workers, Clark said he’s observed senior staff who can dive deep on specific tasks and curious new graduates who “sand off the edges” of AI output showing the clearest value during the AI age. But mid-level engineers may have the toughest time because they may be less inclined to experiment with new AI-driven workflows, he said.

“The future state is less about your ability to be a precise expert on transportation or warehousing or something else, and more about your ability to understand how systems connect and work together,” Clark said. “That makes me hopeful about it because I think that’s much more enjoyable human work.”

breaks the traditional barrier between audience and newsroom. The show transforms

Fortune DailyFortune’s trusted reporting into actionable, conversational, and entertaining insights for an emerging class of business leaders.

Watch here.