Companies have announced 87,714 AI-attributed job cuts through May 2026, according to Challenger, Gray & Christmas—a figure that accounts for 22% of all layoffs this year. It is a story of replacement, where silicon replaces salary. But a new study from the Ramp Economics Lab and Revelio Labs suggests that the reality is far more bifurcated, and perhaps more optimistic, than the headlines suggest.
The study, titled A New Look at AI’s Impact on Jobs: Firm-Level AI Spending and Workforce Adjustment, tracked 21,599 U.S. firms. It found that high-intensity AI adopters—those spending roughly $30 or more per employee per month on AI tools—actually grew their headcount by 10.2% in the two years following adoption. Even more striking, entry-level headcount at these same firms grew by 12%. Rather than a uniform wave of displacement, we are seeing a split: some companies are using AI to expand, while others are using it to trim.
Ara Kharazian, Lead Economist at Ramp, argues that the discrepancy between these findings and the prevailing doom-and-gloom narrative stems from poor data quality in previous research.
The research until now has relied on datasets that are available but not appropriate for these questions, resulting in the general public getting unreliable answers on how AI will actually affect our economy.
Kharazian is blunt about the current discourse, adding:
If you are reading headlines where CEOs blame layoffs on AI, be skeptical.
The study highlights that heavy AI adopters are not random samples; they are typically larger, more engineering-intensive, venture-backed, and faster-growing than their peers. This creates a clear bifurcation. While low-intensity adopters showed no statistically significant change in headcount, the high-intensity group saw a distinct shift, with their entry-level workforce share increasing by 1.15 percentage points relative to non-adopters. The gains, however, are not immediate. The data shows a learning curve, with headcount growth typically taking six to 12 months to materialize after the initial investment.
Yet, the study is not a victory lap for AI proponents. The authors are careful to note that the research is correlational, not causal. They do not claim that spending money on AI tools mechanically forces a company to hire more people. This limitation is where critics find room for doubt. Paul Roetzer of SmarterX points to the missing piece of the puzzle: the counterfactual.
If we don’t have the answer to that question, then this research is basically meaningless.
Without knowing what hiring would have looked like at these firms had they not adopted AI, it is difficult to isolate the technology’s specific impact from the company’s existing growth trajectory.
This tension between the Challenger layoff data and the Ramp-Revelio findings highlights a reality of the current economy: AI is not a singular force acting upon the workforce. It is a tool whose impact is mediated by the strategy, funding, and culture of the firm wielding it. Some companies are using AI to automate away roles, while others are using it to scale operations and lower the barrier to entry for junior talent.
For the average worker, the takeaway is clear: the threat of AI is not distributed evenly. If you are looking at your own career, the sector matters less than the specific company’s approach to technology. As Kharazian noted,
Who funded you is a better predictor of AI adoption than the sector you’re in.
The next time you see a headline about AI-driven job cuts, remember that it is only half the story. The other half is happening in the offices of companies that are using the same technology to grow, not shrink.