The Companies Betting Hardest on AI Are Also the Ones Hiring the Fastest

There is a story everyone already believes about artificial intelligence and jobs: AI comes in, headcount goes down. It is a clean narrative, and it shows up in nearly every boardroom conversation and LinkedIn post about the technology right now. But two new datasets, released within days of each other, tell a more complex story: at the companies committing most heavily to AI, hiring is accelerating, not shrinking.

What the data actually shows

On June 30, Ramp and Revelio Labs published a working paper that tracked AI spending against workforce records for 21,559 U.S. companies from 2021 through early 2026. The researchers split companies into two groups: high intensity adopters, defined as firms spending at least $30 per employee per month on AI tools for three consecutive months, and low intensity adopters, firms making smaller or one time purchases.

The high intensity group grew headcount 10.2 percent over the two years following adoption. Entry level headcount in that same group grew 12 percent, faster than any other role category the researchers tracked. Low intensity adopters saw no statistically significant change in headcount at all.

The timing matters too. Ramp found the hiring effect did not appear until six to 12 months after adoption began. This is not an instant payoff. It shows up only after a team has had time to actually restructure how it works around the tools.

PwC's 2026 Global AI Jobs Barometer, which analyzed more than a billion job postings across 27 countries, found a similar pattern at a much larger scale. Headcount growth at the most AI exposed companies reached 52 percent relative to a 2018 baseline, compared to 36 percent at the least AI exposed companies. That is a gap of 16 percentage points (52 minus 36). Wage growth followed the same pattern, running higher at the companies leaning hardest into the technology.

A caveat

Both research teams describe this as correlation, not proof of causation. Companies that go deep on AI tend to already be larger, faster growing, more technical, and more likely to be venture backed before they ever increase their AI spend. It is possible these companies were already positioned to expand across every function at once, hiring included. The direct causal link between AI adoption and hiring growth has not been established, and both research teams say so explicitly.

Why the mechanism matters more than the headline

That limitation is worth taking seriously rather than smoothing over for a cleaner story. Still, the mechanism both papers point toward lines up with a shift a lot of AI native venture builders, Misfit Labs included, are watching closely: AI does not eliminate the need for people; it eliminates the fixed cost of doing more things.

A team that used to need six people to ship a feature might now need two engineers and a well built AI workflow around them. That does not necessarily shrink the company, but it does free up the other four people to work on things the company never had bandwidth for before: better support, wider sales coverage, faster iteration on the next product. All of that still requires people, it just requires them in a different role or capacity.

The entry level data supports this interpretation. Junior hiring did not contract at the companies leaning hardest into AI. It grew faster than any other hiring category in the dataset. The routine tasks that used to make up most of a junior role are increasingly handled by AI, which means junior hires get pulled into higher judgment work sooner. That is a harder job. It is also a more valuable one, and it changes what companies should be training their youngest hires to do.

The real story is bigger than headcount

There is a point buried under the topline numbers that matters more than the numbers themselves. In its own analysis, PwC found that the companies posting the largest productivity gains were not using AI mainly to cut costs. They were using it to pursue new revenue, enter new markets, and build products that were not possible before. Cost cutting is the easy story to tell about AI. Growth is the harder one to build toward, and it is the one this data says is actually happening at the companies doing this well.

That distinction is worth sitting with. Two years into a very loudest AI adoption cycle, the companies pulling ahead are not the ones that used AI to run leaner. They are the ones that used it to become capable of more, and then hired the people needed to build that more. If the dominant narrative has trained people to hear "AI" and think "layoffs," this data is a useful corrective to add nuance to the discussion. The bigger opportunity is not doing the same work with fewer people; it is doing work that was never possible before, at a scale that requires more people to pull off, not fewer.