Job-postings data rarely settles arguments, but sometimes it shows exactly where the ground is moving. A Draup analysis of Fortune 500 hiring, released 30 September and built on more than a billion job descriptions, finds that AI “builder” roles — AI and ML engineers, generative-AI engineers, full-stack AI developers, AI architects — now make up 27 percent of technology job postings at the largest US companies, more than double their share in 2021. The same data shows AI skill requirements spreading into roles that never used to mention them.
🔍 THE BOTTOM LINE: The composition of corporate tech hiring has changed more in five years than in the previous fifteen: a job category that barely existed in 2021 now accounts for more than a quarter of big-company tech demand, and AI literacy is now asked for in a fifth of HR postings and a quarter of sales ads. For job seekers, the signal is less “learn to code” than “learn to work with AI in whatever you already do.”
What the 27 percent actually measures
Draup’s Macro Labour Market Trends databook splits AI-era technology hiring into six role families. In 2026, AI Builders lead at 27 percent, “forward deployed engineer” roles — the go-to-your-customer deployment jobs — have grown to 18.5 percent, infrastructure has slipped to 20.5 percent, and tech support has fallen to 7.5 percent. Total Fortune 500 postings sit near 500,000, about 2.5 times their 2021 level.
The reshuffle matters as much as the growth. Support- and experience-heavy roles are losing share precisely because AI can absorb that work — while the roles that build and connect AI systems expand. It is the same rotation pattern the TrueUp openings data showed for the wider tech sector this week: churn inside the sector, cuts in some families alongside rapid hiring in others.
AI skills are leaving the IT department
The most consequential finding for non-engineers: AI-skill requirements have reached 31 percent of support postings, 25 percent of sales, 21 percent of finance and 20 percent of HR — off tiny 2020 bases, but rising fast. Soft-skill mentions per posting rose 8.5 points in twelve months, led by “AI Literacy” (up 725 percent year on year) and “Precise Inference” (up 590 percent).
Employers are also rebuilding how people enter. Internships and contract roles now account for 27 percent of early-career hiring, up from 13 percent in 2020 — and they last far longer, with the average internship stretching from 3.6 to 9.7 months. That is closer to a trial employment period than a summer programme. Mid-career employment is growing fastest: up 11.4 percent for workers aged 35–40 since generative AI went mainstream in late 2022, while employment for 22–25-year-olds has fallen 2.4 percent.
New Zealand readers will recognise both patterns. NZ job ads asking for AI skills have more than doubled over the year, and graduates here describe the same tightening entry funnel Draup’s numbers quantify. The age-cohort finding also mirrors local survey work: it is the experienced mid-career cohort, not the newest entrants, capturing most of the AI-era employment growth — consistent with NZ research showing workers gaining most when AI skills stack on top of existing domain expertise.
The hiring geography is shifting too
One quieter datapoint: global tech postings fell 17 percent overall, but the Gulf states surged — Saudi Arabia up 141 percent, Qatar 137 percent, the UAE 101 percent — alongside India (+45 percent), Mexico (+48 percent) and the Philippines (+51 percent). AI hiring is globalising faster than tech hiring generally, which matters for Kiwi workers competing in a remote-first market and for employers here trying to hold senior AI talent against well-funded overseas offers. Stanford researchers flagged the same US–china talent race dynamics this week — the competition for AI skills is increasingly international, and small markets feel it first in salaries.
Cautions before treating this as a map of the future
Draup sells workforce analytics, and its report is drawn from postings at large US corporates — a lens that may not transfer cleanly to small-market labour forces like New Zealand’s, where mid-size firms dominate. Job-posting language also shifts faster than actual hiring: demanding “AI literacy” in an ad costs nothing. And the role-family categories (“builder”, “guardrail”, “experience”) are Draup’s own taxonomy, built on its own classification of more than a billion posts — credible, but not independently audited.
🔍 THE BOTTOM LINE: Postings data is imperfect, but a doubling of the AI-builder share and AI requirements spreading into a fifth of HR ads is the kind of change that shows up in pay packets a few years later. The durable takeaway for careers: the premium is migrating toward people who combine AI fluency with domain judgment — and, at least so far, experience is amplifying that premium rather than erasing it.
❓ FAQ
What is an “AI builder” role? Draup’s term for jobs centred on building AI systems — AI/ML engineers, generative-AI engineers, full-stack AI developers, AI research scientists and AI architects.
Why are internships getting longer? Companies appear to be using extended internships and contracts as trial hires before committing to permanent roles — a shift toward experiential entry points as AI changes what new hires need to prove.
Does this data include New Zealand? No. It covers Fortune 500 hiring, mainly US-based. The trends rhyme with NZ data — SEEK’s NZ reports show AI mentions in job ads up sharply — but the specific percentages are American.
Which skills are rising fastest in job ads? AI Literacy (+725 percent YoY) and Precise Inference (+590 percent) lead Draup’s soft-skill list, inside a broader 8.5-point rise in soft-skill mentions per posting over twelve months.
📰 Sources
- PR Newswire — Draup report announcement, 30 September 2026
- Draup — Macro Labor Market Trends 2026
- ODSC — The Week in AI Careers, 5 October 2026
— CJ Murden, editor of Singularity.Kiwi. Former digital technologies teacher, author of AI-focused books. Writing with a New Zealand focus.
