The most-quoted line in labour economics for the past three years has been some version of “AI is about to take your job.” On September 4, The Economist ran a different headline: “The jobs apocalypse is postponed. An AI jobs boom is here.” The magazine’s own analysis estimates that AI has so far created around 1 million new jobs in America — comfortably exceeding the roughly 200,000 layoffs it links to AI over the same period.
It is the first time a mainstream publication has put a positive net number on AI’s employment effect using hiring data rather than forecasts. Whether the number survives scrutiny is exactly the kind of question worth asking.
What The Economist actually found
The analysis, published alongside a US jobs report that beat expectations, tracks hiring in the industries AI is directly powering and compares it with the job cuts attributed to the technology. Its claim, summarised in the magazine’s own social posts and coverage of the piece: AI-related hiring — from data-centre construction to tech roles — has generated an estimated 1 million positions, while AI-linked job losses sit near 200,000. In a US labour market where employers churn through roughly 1.7 million separations in a typical month, the magazine notes that even headline-grabbing AI layoff announcements get lost in the churn.
The piece also points to occupations once considered most vulnerable to AI now benefiting from AI-era demand. Between 2023 and 2025, employment in some of those categories rose rather than fell — a pattern The Economist credits partly to workers shifting into roles that build, power, or supervise the technology. Its tracking of five industries at the heart of the data-centre build-out, from electrical contracting to equipment manufacturing, shows sustained employment growth as hyperscalers pour hundreds of billions of dollars into physical infrastructure.
That infrastructure angle is not speculative. The New York Times reported in July that AI companies are recruiting electricians and carpenters “by the thousands” — a jobs pipeline that simply did not exist in 2023.
The jobs report that backs the boom case
The timing was not accidental. On the same day, the US Bureau of Labor Statistics reported that employers added 162,000 jobs in August — more than double the roughly 65,000 economists had forecast, with the unemployment rate steady at 4.1 percent. June and July figures were revised up by a combined 55,000. Navy Federal Credit Union chief economist Heather Long called it “a huge August jobs report,” and noted solid hiring in manufacturing and construction, which she linked in part to AI-related data-centre construction.
The 2026 pattern is striking when set against 2025: monthly job growth has averaged roughly 80,000 this year against fewer than 10,000 a month last year, according to an analysis by Hubbard-O’Brien Economics. Whatever is happening with AI, the feared collapse in total hiring has not arrived.
The counterpoint: the cuts data tells a narrower story
The Challenger, Gray & Christmas outplacement report, released September 3, adds texture rather than contradiction. US employers announced 52,881 job cuts in August — up 58 percent on July but down 38 percent year on year, and the slowest August since 2022. Companies blamed AI for just 3,462 cuts in the month, the lowest monthly total since December 2025, and restructuring — not AI — was the most-cited reason for the first time in five months.
For the year to date, though, the picture is less rosy: AI has been cited in 116,175 cuts, about 22 percent of all announced layoffs and the leading single reason for job cuts in 2026. The technology sector’s 155,126 year-to-date cuts remain the largest of any industry. Both things are true at once: AI-linked layoffs are falling month to month, and AI remains the most-quoted justification for cuts in 2025-26’s reorganisation wave.
The occupational detail in the BLS release also gives pause. Unemployment among “management, professional, and related occupations” ticked up year on year — from 2.7 to 2.9 percent — even as sales, office and production occupations improved. The information sector, which covers software and data processing, cut 23,000 jobs in August. As HRD America’s analysis of the same report put it, those numbers are “consistent with an AI-driven story” of deferred hiring and absorbed tasks — but also with higher interest rates and a post-pandemic hiring digestion, and one month of occupational data is not a trend line.
What it means for New Zealand
New Zealand’s labour market runs on a different clock. SEEK’s July report showed NZ job ads falling for a third consecutive month, with unemployment at a 11-year high of 5.6 percent in the June quarter. The AI-driven construction boom The Economist describes has a domestic echo — 173 data-centre jobs were listed on SEEK nationally in early September, and PwC’s NZ AI Jobs Barometer reported AI-related job ads nearly doubling in a year — but a small economy with soft aggregate demand is unlikely to reproduce the US pattern soon.
The transferable insight is directional, not numerical. If AI’s first-order employment effect is to move workers into building, powering and supervising AI systems — electricians, technicians, governance specialists — then the skills question for NZ workers is not “will AI take my job” but “which side of the build-out am I on.” New Zealand’s own data-centre and grid investment, plus the fastest-growing AI skill categories in NZ job ads being ethics and governance roles, suggest at least the early scaffolding of the same shift.
The honest caveat
The 1-million-jobs figure is The Economist’s estimate, not an official statistic, and it counts AI-era demand broadly — everyone from data-centre electricians to newly hired AI product managers. Critics make a fair point: a booming build-out can coexist with a hollowing-out of entry-level pathways, which is exactly what Stanford Digital Economy Lab’s research on junior hiring and New York Fed survey data on reduced entry hiring suggest. The US jobs numbers are also unusually revision-prone this year, and one strong August does not settle anything.
But the direction of the evidence in 2026 has been consistent: fewer AI-attributed layoffs than feared, faster hiring in AI-adjacent industries, and a labour market that keeps outperforming the apocalypse forecasts. For now, the burden of proof sits with the doom scenarios.
FAQ
How many jobs has AI created in the United States? The Economist’s September 2026 analysis estimates AI has so far created around 1 million new jobs in America, exceeding the roughly 200,000 layoffs it links to AI — an estimate based on hiring in AI-related industries, not an official statistic.
How many US layoffs were blamed on AI in August 2026? Challenger, Gray & Christmas recorded 3,462 job cuts attributed to AI in August 2026, the lowest monthly total since December 2025. Year to date, AI has been cited in 116,175 cuts, about 22 percent of announced layoffs.
Did the US add jobs in August 2026? Yes — the BLS reported 162,000 jobs added in August 2026, more than double the forecast, with unemployment steady at 4.1 percent and prior months revised up by 55,000.
Is the AI jobs boom reaching New Zealand? Partially. NZ’s labour market remains soft overall — unemployment was 5.6 percent in the June quarter — but data-centre hiring and AI-skills demand are growing, with SEEK listing 173 data-centre roles nationally in September 2026.
Related reading: Only 4 Percent of Firms Using AI Actually Laid Anyone Off, New Zealand’s AI Hiring Has More Than Doubled in a Year, and The 62 Percent AI Wage Premium Is Now the Labour Market’s Biggest Number.
— CJ Murden, editor of Singularity.Kiwi. Former digital technologies teacher, author of AI-focused books. Writing with a New Zealand focus.
