An AI job growth study published 30 June 2026 by Ramp and Revelio Labs has found that companies spending the most on AI are adding staff faster than their peers, including in the entry-level roles everyone is convinced are already gone. It is a more optimistic headline than the current news cycle deserves, and the authors are the first to say the fine print matters.
What the AI Job Growth Study Actually Measured
The paper matched enterprise AI spending data from Ramp against workforce records from Revelio Labs, covering 21,559 U.S. firms, the snippet characterised this as “nearly 22,000,” but the paper’s own figure is 21,559. The authors are Ara Kharazian, Lisa Simon, and Ryan Stevens.
“High-intensity adopters”, firms spending an average of $30 per employee per month on AI in their first three months, saw headcount rise 10.2% over the following 24 months. That growth ran across engineering, sales, administration, customer service, finance, marketing, and science functions. Low-intensity adopters saw no detectably higher headcount at all.
The entry-level picture is the part that cuts against the prevailing panic. At high-intensity adopters, junior headcount rose 12%, and entry-level workers’ share of the workforce climbed roughly 1.15 percentage points relative to the control group. The strongest gains were in the information sector: software, internet, media, and tech-adjacent firms.
The paper’s own explanation is worth quoting directly: ‘For software and technology firms, AI can make core output cheaper or faster to produce: writing code, debugging, building internal tools, producing technical documentation, and supporting product development. Lower production costs in these workflows can raise the return to expanding the whole firm, not just the engineering team.’
In other words, when AI cuts the cost of producing something, it can make the whole operation worth scaling, rather than simply trimming the headcount that used to produce it by hand.
The Grimmer Context This Data Sits Inside
None of this lands in a vacuum. According to the Challenger, Gray & Christmas May 2026 report, employers cited AI as the reason for 87,714 job cuts in the year to date through May 2026, with 38,579 of those coming in May alone, making AI the leading stated cause of cuts for the third consecutive month.
Total U.S. job cuts in May reached 97,006, up 16% from April and the highest May figure since the pandemic month of May 2020. The cumulative toll through May 2026: 397,755 cuts announced.
Goldman Sachs economists, as reported by Fortune, estimate AI has erased around 16,000 net jobs per month over the past year, with Gen Z and entry-level workers absorbing the heaviest share. Data centre construction (the physical backbone of AI) has added 212,000 jobs since 2022 and is generating roughly 9,000 new positions per month, which softens the net figure but does not erase it.
Goldman Sachs Research puts 300 million jobs globally in the exposure zone for automation, while estimating that if current AI use cases expanded across the whole economy, approximately 2.5% of U.S. employment would face actual displacement risk. A separate Goldman analysis finds that where AI augments rather than replaces workers, monthly payroll growth has picked up; where it substitutes, employment falls. The transition, Goldman projects, could bump the U.S. unemployment rate by approximately half a percentage point while displaced workers find new footing.
The Ramp/Revelio paper does not dispute this macro picture. Its authors are explicit: ‘This paper does not show that AI universally creates jobs, but it does counter claims that AI will lead to broad job losses.’
The Gap That Could Widen
There is a selection problem the data cannot quite escape. The firms logging the strongest hiring gains are tech-forward, often venture-backed, and already on a growth trajectory. As 247 Wall St. notes in its coverage of the study, hiring gains do not materialise until six to twelve months after adoption, meaning quarterly ROI scorecards consistently undercount AI’s workforce impact. But it also means firms that buy subscriptions and run pilots without sustained investment see no headcount gains at all.
The authors put it plainly: ‘Firms without those channels may fall behind.’ The channels in question are capital, technical staff, founder networks, and management bandwidth, the infrastructure of firms that were already well-positioned before they opened their first AI subscription.
That is the uncomfortable shape of the data. AI spending correlates with hiring growth, but the firms doing the spending are not a representative cross-section of the economy. The question of whether AI is broadly a job creator or a job destroyer may depend less on the technology itself than on which kind of firm is holding it.
