LinkedIn’s COO Dan Shapero has a number that cuts against the layoff headlines: 1.3 million. That is how many new AI-enabled roles the platform counts as added globally over the past two years — AI Engineers, Forward-Deployed Engineers, Data Annotators, and titles that did not exist three years ago. On top of that sit more than 600,000 new data centre jobs created in a single year, according to LinkedIn’s Labor Market Report, Building a Future of Work That Works.
The report’s framing is that the global labour market is not retreating so much as rotating. Overall hiring sits roughly 20 percent below pre-pandemic levels, but the growth that does exist is concentrating around the AI buildout — a shift LinkedIn calls the “new-collar” era: work that blends technical fluency with distinctly human strengths.
The Job Titles Are Real. The Pathways Are Not Obvious
Three role families dominate the count. Data annotators — the people labelling and quality-checking the data models learn from — have become one of the largest new job categories in the AI economy. AI engineers remain LinkedIn’s fastest-growing job title in the US for a second straight year. And forward-deployed engineers sit inside client businesses, translating between what a marketing department needs and what an AI system can actually do.
Two things stand out for anyone planning a career rather than a headline.
First, the infrastructure jobs are overwhelmingly practical. Data centre technicians, facilities specialists, network engineers — the 600,000-job buildout is closer to an electricians-and-cooling story than a PhD story. LinkedIn cites US Bureau of Labor Statistics projections that about 60 percent of new jobs by 2030 will come from occupations that do not typically require a degree.
Second, the management layer is being hired too. LinkedIn counted double-digit increases in companies adding Head of AI roles across Australia (up 32 percent), Canada (31 percent), India (30 percent), Germany (30 percent), the UK (30 percent) and the US (28 percent) in a single year. AI governance is moving from committee conversation to org chart.
The New Zealand Read
New Zealand’s own AI strategy, released by MBIE in July 2025, identified lack of AI skills as a core barrier to adoption, citing Datacom survey data that 43 percent of non-adopting businesses name lack of expertise as their main reason for holding back. The strategy’s answer leans on the tertiary system — the University of Auckland’s software engineering degree, Victoria University of Wellington’s Master of AI, plus Budget 2025 funding of $213 million for tuition and training subsidies and $64 million for STEM priority areas.
The LinkedIn data suggests a question that strategy does not fully answer: what if a large share of the new jobs sit outside the degree pipeline? New Zealand’s economy is built on small businesses, agriculture and trades — exactly the base a “smart adopter” country needs to reach. The new-collar pattern implies technician, integrator and governance-adjacent roles that can be trained in months, not years. Whether NZ polytechnics and industry training providers gear up for that demand is arguably a bigger career question for 2027 school leavers than anything happening at Silicon Valley.
There is also a warning embedded in the good news. The same LinkedIn report shows hiring down about 20 percent on pre-pandemic levels and 52 percent of workers job hunting while nearly 80 percent feel unprepared for the search. New job categories do not automatically absorb displaced workers — and this site has covered the Stanford research showing young workers in AI-exposed roles falling further behind. The rotation is real, but catching it requires training that arrives before the jobs do, not after.
FAQ
What are “new-collar” jobs? LinkedIn’s term for roles blending technical skills with human strengths — AI engineers, data annotators, forward-deployed engineers and data centre technicians. Most did not exist three years ago.
How many jobs has AI created so far? LinkedIn counts more than 1.3 million new AI-enabled roles globally over two years, plus over 600,000 data centre jobs in the past year.
Do AI jobs require a university degree? Many do not. LinkedIn cites US Bureau of Labor Statistics projections that about 60 percent of new jobs by 2030 will come from occupations that typically do not require a degree.
Related reading: New Zealand’s AI Hiring Has More Than Doubled in a Year, NZ Employers Want AI Skills More Than Australian Ones Do, and Stanford’s AI Employment Gap for Young Workers Widens to 19 Percent.
— CJ Murden, editor of Singularity.Kiwi. Former digital technologies teacher, author of AI-focused books. Writing with a New Zealand focus.
