A Goldman Sachs Research report released on 21 September offers one of the clearest maps yet of where AI’s labour-market impact is actually landing — and where it isn’t. Economists Sarah Dong and Joseph Briggs, writing in a report titled “Is AI Impacting Global Labor Markets?”, find hiring weakness concentrated almost entirely in white-collar service industries across developed economies, while blue-collar employment has so far been largely spared.
🔍 THE BOTTOM LINE
The AI hiring slowdown is real but remarkably narrow: it hits call centres, software publishing, consulting and advertising — the industries where labour-automating tools are already available. Employment in those sectors has fallen sharply below trend across the developed world, most dramatically in the United States. Meanwhile, economy-wide displacement remains low, and in New Zealand the same white-collar categories are still posting modest job-ad growth. The map matters more than the panic.
What the Goldman Sachs data shows
The report’s headline numbers describe a hollowing-out below the waterline, visible in employment trends rather than mass layoffs. According to the report, call-centre employment now stands 39 percent below trend in the US, 33 percent below in Canada, and 27 percent below in Germany. Software publishing, management consulting and advertising services show similar contractions relative to historical benchmarks.
The economists write that employment in information and communication services — among the most AI-exposed industries — “has slowed across nearly all major developed markets since 2022.” The clearest negative effects are in the US, where tech-sector employment has fallen below long-term trends. Dong and Briggs conclude that the patterns “confirm that, like in the US, AI-employment headwinds are visible in industries where labor-automating tools are already available.”
Two caveats sit right in the report’s own framing. The headwinds are “global in nature,” but outright negative employment impacts are “most compelling in the US” — other developed markets show slowdowns, not collapse. And economy-wide displacement, the report notes, remains low.
The same week, a different reading of the numbers
The Goldman report landed three days after another data point complicated the doom reading. Challenger, Gray & Christmas data showed US employers announced 54,064 job cuts in September, down 37 percent from August — with AI explicitly cited in about 7,000 of them, part of roughly 17,000 for the year so far. That is a real and fast-growing attribution category, but still a small slice of the roughly 946,000 cuts announced in 2026 overall, most of which companies tied to other causes.
Meanwhile, Indeed Hiring Lab analysis published on 17 September found the fastest-rising advertised pay in America is in precisely the occupations most exposed to generative AI — 46 percent growth since 2021, against 25 percent for the least-exposed roles. The two datasets are not contradictory so much as they are measuring different things: pay for people who stay in AI-adjacent work is rising fast, even as aggregate hiring in the most exposed sectors cools.
What the NZ numbers say
New Zealand sits on the gentler end of every curve in this story. SEEK’s August employment report shows NZ job ads up 0.6 percent month on month and 9.7 percent year on year — the 21st consecutive month of growth. Even the white-collar categories Goldman flags as most exposed are still growing here: SEEK’s data shows AI skills mentioned in 9.8 percent of marketing and communications ads and 9.3 percent of consulting and strategy roles, both up on July.
The caution flags are visible too. Stats NZ put unemployment at 5.6 percent in the June quarter, the highest in over a decade, and SEEK notes ads for high-automation-exposure occupations — keyboard operators, telemarketers, HR clerks — dipped 0.1 percent year on year. SEEK itself attributes that only tentatively to AI. The Reserve Bank’s own research, cited by recruiters, found professional, managerial and administrative occupations carry relatively high AI exposure. Goldman’s developed-market pattern — white-collar hiring cooling while physical-work sectors grow — is a plausible preview for NZ if adoption follows the same path, but nothing in the local data yet confirms it has arrived.
FAQ
Which jobs is AI affecting most, according to Goldman Sachs? The report finds the clearest hiring slowdowns in call centres (39 percent below trend in the US), software publishing, management consulting and advertising services — white-collar sectors where automation tools already exist. Blue-collar employment shows far less impact so far.
Is AI causing mass unemployment? Not in the aggregate data Goldman reviewed. Economy-wide displacement remains low, and the impact concentrates in specific service industries rather than spreading across the labour market.
What about New Zealand? NZ job ads rose 9.7 percent year on year in August, and even AI-exposed white-collar categories are still growing locally. Unemployment of 5.6 percent reflects the broader economic cycle, with AI exposure a factor economists are watching rather than a proven local driver of job losses.
— CJ Murden, editor of Singularity.Kiwi. Former digital technologies teacher, author of AI-focused books. Writing with a New Zealand focus.
