AI isn't taking jobs. Companies that invest in it hire more

2026-07-04

The loudest forecast of the past two years says artificial intelligence will wipe out millions of jobs. The freshest data points the other way: companies that invest seriously in AI grow their headcount by 10.2% within two years of adoption, faster than comparable firms that do not. The figure comes from Ramp Economics Lab’s study “A New Look at AI’s Impact on Jobs”, published on 30 June 2026, the first analysis to match real corporate AI spending against hard employment data. Below I break the numbers down and spell out what they mean for employers and HR teams.

What the data shows

The authors combined two datasets that rarely sit side by side: real company payments for AI tools, meaning corporate card and bill-pay records from the Ramp platform, and monthly employment data for those same firms from Revelio Labs, a platform that collects and analyzes workforce data. The panel covered 21,599 U.S. companies and was observed month by month from January 2021 to February 2026. That matters, because earlier work leaned mostly on surveys and self-reported declarations, whereas here a real expense is set against a real job. In fairness I should add that this is an observational analysis, so it shows co-occurrence rather than hard cause and effect. The results are laid out in the full research paper, and independently of Ramp they were also reported by PRNewswire.

The entire increase is driven by heavy adopters

The data points clearly to one relationship: the rise in headcount is driven solely by firms that adopt AI intensively. Simply buying a few licenses and tinkering with the tools now and then produces no measurable hiring effect. Only companies with high AI spending per employee grow noticeably faster than the rest. In other words, what matters is not the mere presence of AI in a company, but the scale and seriousness of the rollout.

Even junior roles are growing

The result I find most interesting concerns entry-level positions. The popular thesis, which I find persuasive too, runs like this: AI will hit junior roles first, because their tasks are the easiest to automate. Ramp’s data shows the opposite direction, because at the companies investing most in AI, entry-level headcount rose by 12% within two years of adoption. A serious adopter does not lay juniors off, it hires them, because a growing company needs people at every level.

Small firms: less often, but more intensively

A separate thread concerns small firms. They adopt AI less often than the big players, but once they decide to, they do it more intensively and with a larger effect. It makes sense to me, because in a small organization a good AI tool unlocks capabilities that previously required hiring a separate team, so instead of cutting jobs the firm gains room to grow. I am, at least in a sense, an example of such a company myself. Even though my business keeps growing, instead of adding headcount I add AI agents.

What this means for employers and HR

So what follows for an employer, an HR manager and a business owner? If a serious AI rollout goes hand in hand with more hiring rather than less, then the edge stops being the mere ownership of tools and becomes the speed and quality of hiring. A company that grows headcount by 10% in two years pushes far more candidates through its recruitment process, so a well-organized process starts to genuinely decide the outcome. That is exactly why I spend my days building the recruitment system Element, which keeps that process in order. The paradox is that the more seriously a company moves into AI, the more human recruitment it has to handle, not less.

What the data does not say

To close, honestly about the limits of these conclusions. The study covers U.S. firms, rests on correlation, and the effect shows up mainly among real, intensive users, so it is not proof that every AI rollout automatically creates jobs. Ramp itself stresses that we are talking about companies that genuinely invest in AI, not the whole economy. So I would not be honest if I turned this data into a slogan that AI always means more work. The report’s data says something narrower but important: among firms that treat AI seriously, fears of mass layoffs have not materialized so far. In my view this comes down to two things: on one side the potential of AI keeps growing and opens new avenues for people and organizations to develop, and on the other AI is still not a mature enough technology to do work at a level close to an average specialist.