Heavy spenders on artificial intelligence don’t slash payrolls. They expand them. Firms that pour serious money into the technology see their workforce swell by more than 10 percent in the two years after they dive in. Low-intensity users register no meaningful shift at all.
The pattern comes from a fresh analysis of spending records from more than 21,000 U.S. companies. Ramp, the corporate card and finance platform, teamed with workforce data provider Revelio Labs. They linked actual AI-related expenditures to headcount figures. The result challenges the tidy story that artificial intelligence simply replaces humans.
High-intensity adopters, those spending roughly $34 per employee per month in the early phase, drive all the growth. Entry-level roles expand even faster, up 12 percent. Sales, engineering, administration and customer service positions all rise together. The gains arrive gradually, often six to 12 months after the first big outlays. Best practices take time to spread, the researchers argue.
But the data carries a caution. Companies that adopt early already tend to grow faster, attract venture capital and employ more engineers. The study controls for those differences by comparing early and later adopters within the same intensity group. Still, Ara Kharazian, lead economist at Ramp, flagged the need for skepticism in a social media post. “This is our first evidence that high-AI-adopting firms are hiring different kinds of employees,” he wrote. “We believe they are selecting for a new set of skills, specifically, people who know how to use AI and use it well. Entry-level workers, especially recent graduates and college students, are a natural place to look.”
The findings echo a July 2 report from The Register. Senior reporter Thomas Claburn highlighted the same Ramp study and noted the lag before job gains appear. He also pointed to countervailing pressures. Oracle, for instance, booked roughly $86,000 in severance and restructuring costs for each of the 21,000 workers it cut last year even as it ramped up AI capital spending.
Broader surveys paint a similar mixed picture. Workers in organizations using AI report more turbulence. Gallup found that 34 percent of employees at AI-adopting firms said their employer was hiring and expanding the workforce. Only 28 percent of those at non-adopting organizations said the same. Reductions also ran higher, 23 percent versus 16 percent. Net effect still tilted toward growth, yet the changes felt more pronounced where AI had taken root. Disruption hit 27 percent of workers in adopting organizations compared with 17 percent elsewhere.
New evidence from recent weeks reinforces the hiring side of the ledger.
A Financial Times analysis published days ago reached the same core conclusion: companies that spend more on AI also increase worker numbers faster than peers. The pattern holds across sectors. Meanwhile Boston Consulting Group projected in April that AI would reshape 50 to 55 percent of U.S. jobs in the next two to three years, with augmentation and new roles outpacing outright elimination. Full substitution might affect only 10 to 15 percent of positions five years out.
Productivity gains explain part of the expansion. When AI lifts output, demand for the underlying product or service often rises too. More sales require more people to handle complex tasks, manage exceptions, or pursue new opportunities the technology opens. MIT Sloan researchers reached a parallel finding last year. Firms that adopted AI aggressively recorded faster revenue and employment growth over five years. The productivity boost let them make more while keeping or adding staff.
Yet not every organization sees the upside. Some executives still chase headcount reductions. Surveys from late 2025 showed nearly three in 10 companies had already replaced some roles with AI, with another 37 percent expecting to do so by the end of 2026. Those moves often target specific tasks rather than entire jobs. A U.S. Census Bureau survey found 27 percent of AI-using firms replaced worker tasks, but only 5 percent reported employment changes. Expectations point to 35 percent task replacement and 12 percent employment impact in coming years, with increases slightly more common than cuts.
The disconnect shows up in entry-level markets. Recent college graduates faced a 5.6 percent unemployment rate in March 2026, above the 4.3 percent overall figure at the time. The national rate held near 4.2 percent in June. Kharazian’s hope that AI-savvy juniors would find favor has not yet lifted the broader cohort. Many companies still hesitate to hand critical work to unproven talent, no matter how fluent they are with large language models.
Trust issues complicate the picture further. Palantir CEO Alex Karp told CNBC that both military and commercial customers want control over their compute, models, data and investment returns. They resist depending on frontier providers that might change terms, refuse requests or raise prices. “They want to know they own the means of production,” Karp said. His critique lands because many enterprises now spend heavily on AI yet still need humans to oversee outputs, fix hallucinations and maintain accountability.
That oversight creates its own jobs. The Register noted a surge in UK hiring for roles that essentially babysit the bots. PwC data from mid-June showed AI-related vacancies jumped 61 percent even as overall hiring slowed. Demand focused on users of the technology, not just builders. New positions in governance, prompt engineering, model evaluation and agent operations have multiplied. WeCloudData listed seven emerging AI roles organizations chased in 2026, from AI product managers to governance specialists.
So the technology augments more than it automates, at least so far. BCG calls this the amplified role. When productivity rises and demand expands, humans stay central. Wages can climb as companies compete for talent that pairs judgment with machine output. Lawyers advising on strategy, engineers directing complex projects and salespeople closing nuanced deals all fit the mold. Their numbers hold or grow.
Leaders who treat AI as a pure labor substitute often miss the larger opportunity. They freeze hiring, push existing staff harder and redesign nothing fundamental. Gallup observed that while 65 percent of workers in adopting organizations credit AI with productivity gains at the task level, only about 10 percent see organization-wide transformation. The gap signals that many firms have bolted the technology onto old processes instead of rethinking them.
Frontier firms represent the other path. Microsoft research from earlier this year found that companies combining human expertise with AI agents take on more work, report higher confidence and expand capabilities. More than 80 percent of their leaders expect AI-powered labor to help grow the workforce in the next 12 to 18 months. The pattern matches the Ramp data. Serious investment paired with serious integration produces scale that requires people.
Of course risks remain. Anthropic’s own labor market study found no clear rise in unemployment for highly exposed workers since late 2022, but it detected slower hiring for younger people in those occupations. A 14 percent drop in job-finding rates for 22- to 25-year-olds in exposed fields was suggestive though not definitive. The data underscore that even without mass layoffs, opportunity can narrow for those without the right preparation.
Executives therefore face a choice. They can chase short-term cost savings by replacing routine tasks and thinning ranks. Or they can invest in the complementary skills that turn AI from a novelty into a growth engine. The latter group appears to be winning on headcount, revenue and capability. The former may post lower labor expenses for a quarter or two before discovering that output stalled and customers drifted.
The evidence keeps mounting. Heavy AI spenders add staff faster. They hire different people. They reshape roles around human strengths. And they do it while low adopters tread water. The age of artificial intelligence may not eliminate jobs on net. It could instead reward the organizations wise enough to pair silicon with more, better-trained brains.
WebProNews is an iEntry Publication