The “Zero Evidence” Claim About AI and Jobs Deserves a Closer Look

Apollo Global Management’s chief economist, Torsten Sløk, published a blog post on Friday that’s been making the rounds: “zero evidence of job losses because of AI.” He points to ADP employment data showing private companies added nearly 110,000 jobs in April. His argument is that companies are hiring AI implementation experts, the data center buildout is pushing wages up for AI specialists, and the whole spending wave is creating demand, not eliminating it. He calls it Jevons paradox: cheaper technology generates more consumption, not less. “It is Jevons paradox playing out in real time: cheaper technology is creating more demand and more jobs.” Box CEO Aaron Levie, Dell CEO Michael Dell, and White House AI czar David Sacks all backed the take over the weekend. Goldman Sachs CEO David Solomon made a similar case in a New York Times op-ed. An EY survey of 240 financial services CEOs found that about 60% expect AI investment to maintain or grow their headcount in 2026.

Sløk is also reading the ADP data as a forward-looking signal: he thinks May nonfarm payrolls could come in “significantly higher” than the 95,000 consensus expects. We’ll see Friday when the BLS report drops.

The optimistic case has real numbers behind it. ADP added 110,000 private jobs in April. That’s a fact. It’s also what Altman said last week and what Nvidia’s Jensen Huang called “lazy” thinking when applied to every AI layoff announcement. But “zero evidence” requires flattening a lot of variation into a single aggregate, and that aggregate misses the people who are actually getting hurt. At least a dozen major employers cited AI in their 2026 layoff announcements: Block cut from over 10,000 to under 6,000, Cisco, Atlassian, Cloudflare, Coinbase, and IBM among them. The 2026 layoff wave is real and documented. Some of it is genuinely AI-driven. Some is what Altman called “AI washing”: companies using AI as a clean narrative for cuts they’d have made anyway after over-hiring in 2021 and 2022. The difference between those two things matters a lot for how you plan a career or a hiring strategy.

What the Economic Research Actually Shows

MIT Technology Review’s David Rotman pulled together the latest economic research on this question, and the honest answer is more nuanced than either the panic or the reassurance.

The aggregate data isn’t alarming. Analysis of BLS data shows unemployment for jobs most exposed to AI is actually lower than for jobs with less AI exposure. Erika McEntarfer, who ran the Bureau of Labor Statistics until she was fired last fall after a jobs report that displeased the administration, is now a fellow at the Stanford Institute for Economic Policy Research. Her read: “All of the available evidence to date suggests that AI’s impact on current labor market conditions is likely small right now.” She points to Census data showing only one in five U.S. companies use AI in any business function at all. The technology can’t transform labor markets until it first transforms businesses, and that’s still early.

The exception is entry-level. The Stanford Digital Economy Lab’s “Canaries in the Coal Mine” paper found a 16% decline in entry-level positions in AI-exposed occupations for workers aged 22 to 25, while headcount for older workers in the same fields grew. Recent college graduates are already sitting at 5.6% unemployment, well above the overall rate and the worst since the pandemic. A Federal Reserve Board paper found that coding employment growth slowed by about 3% since ChatGPT launched, though overall coding employment is still rising.

The pattern that explains it: where AI does the task with limited human involvement, entry-level jobs shrank. Where AI makes a person faster, headcount grew. That’s the automation-versus-augmentation split, and it’s the most useful lens for thinking about any specific role.

What I see from the staffing side lines up with the research. The companies making smart decisions right now are asking which junior roles can be redesigned around AI tools, not simply eliminated. The ones quietly stopping entry-level hiring will be looking for mid-level talent in three to four years that doesn’t exist because they chose not to train any. That’s not a hypothetical. It’s the seed-corn problem, and it compounds quietly.

Erik Brynjolfsson at Stanford’s Digital Economy Lab made the point worth circling: hundreds of billions are going into deploying AI, and “we’re not investing even 1% of that on understanding the transition.” The question worth asking isn’t whether AI creates or destroys jobs in the aggregate. It’s who specifically gets hurt in the transition and whether anyone with authority to act is paying close enough attention to do something about it.

About Pete Newsome

Pete Newsome is the President of 4 Corner Resources, the staffing and recruiting firm he founded in 2005. 4 Corner is a member of the American Staffing Association and TechServe Alliance and has been Clearly Rated's top-rated staffing company in Central Florida for seven consecutive years. Recent awards and recognition include being named to Forbes' Best Recruiting and Best Temporary Staffing Firms in America, Business Insider's America's Top Recruiting Firms, The Seminole 100, and The Golden 100. Pete is a freqent conference speaker on the topic of AI's impact on jobs, and he hosts Cornering The Job Market, a weekly show covering real-time workforce trends, analyisis, and news. Connect with Pete on LinkedIn