Three employees collaborating at a plywood desk in a sunlit small New Zealand business office while a manager in a bobble beanie reviews a thin stack of CVs, a kiwi-shaped mug visible, warm morning light, documentary editorial photography
Career & Future

Layoffs Sit at 57-Year Lows While Hiring Freezes: Inside the Low-Hire, Low-Fire Labour Market

Nobody is being fired and nobody is being hired: fresh US claims data, September payrolls and Challenger numbers describe a labour market where the safest place is the job you already have.

labour marketlayoffsAI jobshiringemployment data

The scariest thing about the 2026 labour market is not the layoffs. It is how few of them there are — set against how little hiring is happening.

New US data released this week completes a picture that has been assembling for months. Initial jobless claims fell to 197,000 in the week ending 3 October, the fourth straight week below 200,000 — territory, per Reuters reporting carried by Quartz, not seen since the late 1960s. Layoffs, in other words, are near historic lows. Nobody is firing.

But the Department of Labor’s claims report also shows continuing claims climbing to 1.716 million, and September nonfarm payrolls disappointed again: just 29,000 jobs added, per Bloomberg. The median stretch of unemployment has stretched to 11.5 weeks, close to its highest in four and a half years. Nobody is firing, and nobody is really hiring either. Economists have settled on a phrase for it — the “low-hire, low-fire” labour market — and it quietly rewrites the risk calculus for anyone planning a career move.

🔍 THE BOTTOM LINE: US jobless claims at 197,000 mean layoffs are rarer than at almost any point since the late 1960s, but hiring has stalled to 29,000 jobs added in September. For workers with jobs, this is historically unusually safe ground — for job seekers, it is a long queue in front of a slow window. AI is implicated on the “low-hire” side more than the “low-fire” side.

The numbers behind the strange calm

Start with the firing side, because the headline numbers are startling. Initial claims of 197,000 beat the 200,000 forecast, the four-week average fell to 198,000, and claims are running 36,000 below the same week a year ago. Only about 1.1% of covered jobs are seeing benefit claims — the insured unemployment rate, unchanged, and down from 1.3% last year. Tech-sector layoff trackers tell the same story in that industry: roughly 190,000 tech layoffs year-to-date by TrueUp’s count, tracking toward half of the industry’s 2023 peak of about 430,000, per Business Insider.

Challenger, Gray & Christmas’ September report, which we covered when it landed, showed US job cuts falling to a four-year low even as AI was named as a reason in 21% of the cuts announced. AI is being cited often — but the total pile of announced cuts keeps shrinking.

The hiring side is where the market jams. Employers have added an average of about 80,000 jobs a month this year — a genuine pickup from 2025’s roughly 9,700 — but September’s 29,000 shows how fragile that average is. Continuing claims rose 17,000 to 1.716 million in the latest reading, meaning people who lose a job are taking longer to land the next one. Economists attribute the restraint to tariff and geopolitical uncertainty, an ageing workforce, and a shrunken migration pipeline — but one more factor sits closer to this site’s beat.

Where AI actually shows up

Here is the pattern the raw averages hide: companies are not cutting because of AI, and they are not hiring partly because of it.

Workday’s Global Workforce Report, released 5 October, found 40% of business leaders expect AI to help them get more out of the employees they already have, while just 28% expect it to reduce headcount. Read that in this week’s data and the low-hire, low-fire equilibrium acquires a mechanism: if your plan is to extract more output from existing staff — with AI doing the leverage — then both hiring and firing are unnecessary. The workforce you have becomes the workforce you intend to keep, re-tooled rather than traded. That report’s deeper finding is that jobs are being rewritten, not removed, which is exactly what a market with no churn looks like: same people, different tasks.

That logic is kind to insiders and brutal to outsiders. Workday’s own applicant data showed the median filled job now draws 69 applications, up from 58 a year earlier. A labour market can be the safest on record for incumbents and the most discouraging on record for entrants at the same time, and this one is. Our earlier reporting on the two-margin model — AI spenders hiring more, not less, while cautious firms freeze — fits the month shaped by this week’s numbers: the freeze is real, but it is a strategy, not a verdict on AI.

One caveat belongs here: claims data describe the US, and no single country’s flows determine New Zealand’s. But job-ad flows, hiring-hesitancy and AI-adoption narratives cross the Pacific quickly, and NZ’s own data rhymes — see below.

What it means for New Zealand

This week’s US picture should be read beside the most recent New Zealand numbers. Stats NZ put the June 2026 quarter unemployment rate at 5.6%, up from 5.4% in March, with 440,000 people underutilised — of whom about 166,500 were unemployed, a fifth of them for more than a year, and annual wage inflation at just 2.0%. That is the same shape as the American market, scaled down: layoffs are not the story here either; the story is slow creation of new roles and a long queue. Canada’s official statisticians reached a related conclusion this month, counting digital-economy workers at one in eight Canadian jobs and growing — evidence that the jobs being created, once created, are durable.

The practical read for NZ workers splits by which side of the desk you are on. If you are employed, the data says your position is unusually secure by historical standards — but internal mobility is the ladder to watch, since external movement is slow everywhere and NZ employers’ AI-skills demand keeps compounding. If you are job-seeking, the numbers argue for patience in expectation-setting and precision in targeting: in a market adding roles slowly, the applications that win are the ones that match the rewritten job descriptions — building-level AI skills, per Workday’s skills data — not volume applications into a 69-applicant queue.

For employers, one uncomfortable mirror: keeping everyone and hiring no one feels prudent this week, but firms that freeze development of the people they keep are spending their security cushion. The market that stops firing eventually starts hiring, and the employers with re-skilled incumbents will hire least and grow most.

❓ FAQ

What are US jobless claims and why do they matter? Initial jobless claims count new applications for unemployment benefits each week and act as a near-real-time gauge of layoffs. At 197,000 for the week ending 3 October 2026, they are at their lowest sustained level since the late 1960s — a signal that mass firing is not happening.

What does “low-hire, low-fire” mean for the job market? It describes a labour market where employers are neither cutting staff nor adding many: layoffs sit near historic lows (claims under 200,000 for four straight weeks) while September payrolls added just 29,000 jobs. Incumbent workers are unusually safe; job seekers face a slow market.

Is AI causing the layoffs in this data? AI is being named as a reason in a rising share of announced cuts — 21% in Challenger’s September report — but total announced job cuts fell to a four-year low, and surveys such as Workday’s show 40% of leaders expect AI to get more from existing staff rather than cut headcount. The stronger AI effect is on hiring plans than firing counts.

How does this affect New Zealand workers? NZ’s June-quarter unemployment was 5.6% with 440,000 underutilised, per Stats NZ — the same slow-hiring shape. Employed Kiwis are relatively secure; seekers should target roles squarely and build building-level AI skills that rewritten job descriptions ask for.

🔍 THE BOTTOM LINE

The week’s data describes a labour market in two halves: layoffs near fifty-seven-year lows for those already in jobs, and a hiring window moving at a crawl for those outside. The AI story has migrated from the firing column to the hiring column — leaders plan to leverage the staff they have rather than replace them, which keeps the workforce stable and the queue long. For careers, the strategy writes itself in both countries: insiders, keep building skills while the ground is calm; outsiders, stop spraying applications into queues of 69 and out-skill the queue instead.

📰 Sources

— CJ Murden, editor of Singularity.Kiwi. Former digital technologies teacher, author of AI-focused books. Writing with a New Zealand focus.

Sources: US Department of Labor weekly jobless claims, week ending 3 October 2026 (via Quartz, 8 October 2026, and BabyPips, 9 October 2026), Bloomberg — 'US Firms Add Just 29,000 Jobs, Unemployment Rate Ticks Up' (2 October 2026), Challenger, Gray & Christmas September 2026 job-cut report, as covered on Singularity.Kiwi (2 October 2026), Workday Global Workforce Report (5 October 2026)