Each new round of layoffs blamed on artificial intelligence has fed a simple, ominous story: AI is coming for everyone’s job. Tens of thousands of cuts have already been explicitly tied to automation, and some forecasts suggest that as much as 15% of U.S. roles could disappear over the next five years. For students and early-career workers, it can feel like they are training for jobs that will not exist.

A new analysis from Ramp and Revelio Labs complicates that narrative. By combining data on enterprise AI spending with workforce records from nearly 22,000 companies, the researchers found that firms investing most aggressively in AI are not shrinking their workforces. They are expanding them.

So-called high-intensity adopters, companies spending about 30 dollars per employee per month on AI tools in their first quarter of use, increased headcount by just over 10%. Growth was broad-based, touching engineering, sales, administration, customer service, finance, marketing, and scientist roles. The information sector, which includes software, internet, media, and adjacent tech firms, saw the strongest gains.

That runs counter to the idea that AI is hollowing out junior roles. While other research, including from Goldman Sachs, suggests AI-linked automation has erased thousands of net jobs per month, disproportionately affecting Gen Z and entry-level workers, the Ramp and Revelio report finds that in tech-forward firms, entry-level headcount actually rose by 12%.

The authors are careful not to oversell the findings. Their data skews toward fast-growing, venture-backed, knowledge-work companies that might be hiring anyway. The report does not prove that AI universally creates jobs. It does, however, undercut the claim that AI inevitably triggers broad-based job losses.

One emerging pattern is that AI can function less as a tool for replacing labor and more as a lever for expansion. In software and technology firms, AI can make core tasks cheaper and faster: writing and debugging code, building internal tools, producing documentation, and supporting product development. Lowering those costs can increase the payoff from scaling the entire company, not just the engineering team.

But the benefits are not evenly distributed. Companies that dabble in AI through subscriptions and pilots, without committing to sustained investment and integration, see little change in headcount. That raises the prospect of a widening gap between firms with the capital, technical talent, and leadership capacity to turn AI into growth, and those stuck in perpetual experimentation. In this version of the future, AI does not simply destroy or create jobs; it helps decide which companies, and which workers, get left behind.