Every few weeks a new headline warns that AI is about to hollow out the workforce. So it is worth pausing on what businesses themselves are actually reporting — because the most detailed survey of firm behaviour, published by the Federal Reserve Bank of New York on 1 September, tells a markedly quieter story.
The New York Fed has asked firms in its region about AI adoption and workforce effects every year since 2024. This year’s results, drawn from its August Empire State and business leaders surveys, show adoption has surged: 61 percent of service firms and 51 percent of manufacturers reported using AI, up from 40 percent and 26 percent just a year earlier.
Layoffs remain the exception, not the rule
Despite that jump in adoption, the workforce effects are strikingly small. Only 4 percent of service firms reported laying off workers in response to AI over the past six months, and no manufacturers reported AI-driven layoffs at all this year or last.
The bigger story is retraining. Just over a third of service firms using AI said they retrained existing workers in response to it, along with more than 20 percent of manufacturers. The New York Fed researchers describe retraining as “the primary way firms are adjusting their workforces” — and the training is mostly practical: basic AI literacy, automating routine tasks, prompt engineering, and teaching staff to verify AI outputs rather than over-rely on them.
Hiring effects cut both ways. About 15 percent of service firms said they hired fewer workers than they otherwise would have because of AI, but roughly 13 percent said they hired more workers to help them use it. Those two forces nearly cancel out.
Why the panic and the data disagree
The Fed researchers point to research consistent with their findings — several NBER working papers also find limited labour market effects from AI adoption so far. But they acknowledge one significant counterpoint: the Stanford Digital Economy Lab’s “Canaries in the Coal Mine” research, which suggests entry-level workers may face real barriers as AI takes over the routine tasks junior staff traditionally cut their teeth on.
Both things can be true. Experienced workers appear to be being retrained and augmented far more often than replaced. New entrants are finding the first rung of the ladder harder to reach. The New York Fed’s own data hints at this: the firms scaling back hiring because of AI are, in effect, not backfilling junior roles.
What this means in New Zealand
New Zealand’s labour market skews heavily toward small firms — businesses that, like the Fed’s respondents, typically adopt AI through cheap or free tools rather than major strategic investment. Three-quarters of the Fed’s service firms described their AI spending as minimal to modest, and the median firm had fewer than one in five workers actually using AI. If that pattern holds here, the “AI replaced my whole department” story will stay rare for now, while the “my job quietly changed” story becomes universal.
The practical takeaway from the survey is about who wins the adjustment: firms retraining their existing people, and workers who volunteer for that training early. MBIE’s own modelling, reported by Stuff in July, estimated only 2 percent of current New Zealand occupations face high automation risk at AI’s current level of capability — a reminder that the aggregate data and the anxiety are running on different tracks.
None of this guarantees the picture stays benign. The New York Fed researchers are careful to note these patterns “could shift as adoption matures.” But three years of survey data point one direction so far: businesses are using AI to change work, not to cut it.
FAQ
What did the New York Fed survey find about AI and layoffs? Only 4 percent of service firms using AI reported laying off workers because of it over the past six months, and no manufacturers reported AI-driven layoffs. About a third of service firms retrained workers instead.
How many businesses are using AI? In the New York Fed’s August 2026 survey, 61 percent of service firms and 51 percent of manufacturers reported using AI, up sharply from 40 percent and 26 percent in 2025.
Is AI affecting hiring? Some. About 15 percent of service firms hired fewer people because of AI, but 13 percent hired more to help use it — a near wash overall. The pressure appears concentrated on entry-level roles.
What does this mean for NZ workers? NZ firms mirror the small-firm, low-investment adoption pattern in the survey, suggesting job change rather than job loss is the dominant near-term experience — making retraining the highest-value career move.
— CJ Murden, editor of Singularity.Kiwi. Former digital technologies teacher, author of AI-focused books. Writing with a New Zealand focus.
