The AI infrastructure boom is creating jobs faster than companies can fill them. Kelly’s 2026 Data Center Salary Guide, released August 25, benchmarks AI engineers at a $175,000 national midpoint salary — and warns that 90% of data center operators cite staffing shortages as a critical constraint on growth.
🔍 THE BOTTOM LINE
The same AI wave displacing workers in customer service and back-office roles is generating a hiring frenzy in infrastructure. Permanent data center employment is projected to reach 650,000 positions in the US by 2026, a 30% increase from 2023. Construction-related data center jobs are expected to exceed 180,000 through 2028. But roughly half of 2026 US data center deliveries could face delays or cancellations — not because of power or permits, but because there aren’t enough people to build and run them.
The Numbers
Kelly’s guide combines proprietary workforce data, partner labour-market data, and U.S. Bureau of Labor Statistics wage information to benchmark more than 60 data center lifecycle roles. The headline figures, as reported by Stock Titan:
- AI Engineer: $175,000 national midpoint
- Data Center Operations Leader: $178,000
- Commissioning Program Manager: $135,000
- Construction Manager: $128,000
- Data Center Technician SME: $112,000
The $175K figure for AI engineers aligns with broader market data. Forbes reported in August that AI engineering roles have surged 255% year over year, with average salaries of $113,000 nationally — but up to $250,000 at VC-backed startups. The Kelly midpoint sits higher because data center roles tend to require specialised infrastructure knowledge on top of AI engineering skills.
Why Staffing Is the Bottleneck
The AI buildout has a people problem. Kelly reports that 25% of data center personnel are hired away by competing operators — poaching is the dominant talent strategy because training pipelines can’t keep up. With permanent data center employment projected to hit 650,000 positions (a 30% jump from 2023) and construction jobs exceeding 180,000 through 2028, talent competition has reached what Kelly describes as “critical levels.”
The guide recommends a shift from poaching to pipeline-building: combining targeted hiring with rapid upskilling programs that draw talent from adjacent industries. Workforce acceleration programs, Kelly argues, can build sustainable talent pipelines from sectors like traditional IT, electrical engineering, and construction — reducing reliance on poaching and keeping large AI data center projects on schedule.
What This Means for Job Seekers
For workers thinking about career pivots, the data center salary guide offers a clear signal: the AI economy doesn’t just need software engineers. It needs people who can build, power, cool, and maintain the physical infrastructure that AI runs on.
The fastest-growing roles aren’t all coding jobs. Commissioning program managers, construction managers, and data center technician SMEs all command six-figure salaries. The common thread is specialised knowledge that sits at the intersection of traditional engineering and AI-era infrastructure.
This aligns with DataCamp’s recent analysis of two million job postings, which found that communication skills — not just technical ability — were a universal requirement across all 25 AI and data career paths tracked. The ability to explain technical work to non-technical stakeholders matters as much as Python proficiency. For more on the broader AI engineering hiring boom, see our coverage of Gen Z workers thriving in AI roles.
The NZ Connection
New Zealand’s data center market is smaller but growing. SEEK currently lists 164 data centre jobs across the country, spanning operations, engineering, and facilities roles. CDC Data Centres, which operates across Australia and New Zealand, has achieved NVIDIA DGX-Ready certification — signalling that NZ-based facilities are being built to handle AI workloads.
Globally, Anthropic and OpenAI are both hiring for data center roles in Sydney, including senior data center mechanical engineers specialising in liquid cooling. For NZ workers with infrastructure skills, the trans-Tasman market offers a natural extension — the same staffing shortage driving US salaries is hitting the Asia-Pacific region.
The broader NZ picture is mixed. As Stuff reported, international hiring trends for developers, lawyers, and radiographers have actually increased, with US coder listings up 11% year on year. The jobs apocalypse narrative, while compelling, doesn’t match the hiring data — at least not in the infrastructure layer of the AI stack.
❓ FAQ
What is a data center lifecycle role? A job involved in any stage of a data center’s existence — from site selection and construction through commissioning, operations, maintenance, and decommissioning. Kelly’s guide benchmarks over 60 such roles.
Do I need a computer science degree to work in AI infrastructure? Not necessarily. Construction managers, commissioning program managers, and data center technicians come from electrical, mechanical, and facilities backgrounds. The common requirement is specialised infrastructure knowledge, not software engineering.
Why are AI engineers paid more in data centers than elsewhere? Data center AI engineers typically need infrastructure expertise on top of standard AI engineering skills — knowledge of cooling systems, power distribution, and hardware optimisation. That combination is rare and commands a premium.
Is this hiring boom sustainable? Kelly’s data suggests the shortage will persist through 2028 at minimum, driven by the scale of planned AI infrastructure investment. Whether salaries stay at current levels depends on how quickly training pipelines and upskilling programs can expand the talent pool.
🔍 THE BOTTOM LINE
The AI jobs story has two halves, and the data center salary guide is a reminder that one of them is booming. While 205,000 US workers have been cut in AI-attributed layoffs through August 2026, the infrastructure layer is generating six-figure roles faster than it can fill them. The bottleneck isn’t capital or compute — it’s people. For workers willing to acquire specialised skills at the intersection of engineering and AI infrastructure, the market is wide open. The question is whether training pipelines can scale before the buildout stalls.
