The Federal Reserve Bank of New York just published research that cuts through a lot of the noise around AI and jobs — and the findings are striking.
In their latest Business Leaders Survey, the New York Fed found that more than 61% of service firms are now using AI as part of their business processes. Two years ago, that number was 25%. Among manufacturers, adoption has nearly tripled, from 16% to 51%.
That is an extraordinary pace of change. But here is what makes this research worth paying attention to: almost none of that adoption has translated into layoffs.
The Numbers That Matter
The Fed researchers asked firms directly whether AI had caused them to reduce headcount over the past six months. Only 4% of service firms said yes. Among manufacturers, the number was zero.
That does not mean AI is having no effect on how work gets done. It clearly is. But the effect looks much more like reshaping than replacing. The research found that firms are primarily focused on “helping employees do their current jobs more effectively rather than preparing them for entirely new roles.”
In other words, the typical AI deployment story right now is not: replace the accounts team with a bot. It is: give the accounts team tools that let them handle more work with the same headcount, catch errors faster, and stop doing the repetitive stuff that was eating their week.
The Investment Reality
One finding that stood out: most of this adoption is happening on fairly modest budgets. Three-quarters of service firms and more than 90% of manufacturers describe their AI investments as minimal to modest.
That matters because it tells you where we actually are in the adoption curve. The hype cycle says every company is deep into enterprise AI transformation. The data says most companies have dipped a toe in — a tool here, an automation there — and are watching what happens.
The firms getting real productivity gains are the ones who have gone further. But even they are not burning down headcount. They are choosing to do more with the capacity they freed up.
What This Means for Business
There are two ways to read this research.
The first: AI is not the job-killer the headlines said it would be. That is true, and it is worth saying clearly. If you have employees who are anxious about AI, this data gives you something concrete to share.
The second reading is more important for business leaders: the adoption gap is a competitive problem in the making.
If 61% of your peers in services are now using AI — and that number was 25% just two years ago — the question is not whether to adopt AI. It is how much ground you have already lost by moving slowly, and how quickly you can close it.
The firms that went from zero to meaningful AI use in two years did not do it by buying expensive platforms and hoping for the best. They built capability, trained their teams, and started small on workflows where the ROI was obvious.
That is exactly the pattern Enterprise DNA sees across the businesses we work with. The ones pulling ahead are not spending the most on AI tools. They are the ones who built the internal fluency to actually use them.
The Adoption Surge Is Not Slowing Down
What the Fed data does not show is any sign of the adoption rate cooling off. From 25% to 40% to 61% of service firms in three consecutive years suggests this is still early innings, not a plateau.
That has real implications for workforce planning. The research found that firms are currently focused on making current roles more effective rather than building entirely new AI-specific roles. But that balance will shift. The businesses investing in upskilling their teams now are building the internal capability that will matter when AI adoption gets more sophisticated.
The 75% of firms still at minimal AI investment are not standing still out of choice. They are often held back by the same thing: people who understand the business deeply but lack confidence with the AI tools available to them. Closing that gap is not a technology problem. It is a learning and change management problem.
The Practical Takeaway
The New York Fed’s research lands at a useful moment. It confirms that AI adoption is real and accelerating. It also confirms that the most common outcome is not disruption — it is quiet, incremental improvement that compounds over time.
For business owners, that is both reassuring and a call to action. The disruption is not imminent in the way the headlines suggest. But the compounding advantage that comes from building AI capability into your team is real, and the firms that started two years ago are already pulling ahead.
The window to catch up is still open. But it is getting smaller every quarter.
Enterprise DNA works with businesses at every stage of AI adoption — from teams just getting started with data and AI tools through EDNA Learn, to companies ready to deploy AI agent workforces that handle real operational work. If you want to understand where your business sits in this adoption curve and what the practical next steps look like, book a discovery call.