4 min read
ARTIFICIAL INTELLIGENCE

Hidden fees starve A/NZ innovation budgets

Heather Wright · November 19, 2026
A strip torn through plain grey paper revealing an extreme close-up of the eyes and hair of an engraved banknote portrait

Unexpected charges stall hiring, R&D and AI projects…

Local businesses are among the most likely in the developed world to exceed their tech infrastructure budgets, with new research suggesting difficult-to-forecast provider charges and hidden fees are emerging as a major contributor to cost overruns – and those overspends are forcing companies to scale back on hiring, R&D and AI.

The Wasabi-commissioned study by the Centre for Economics and Business Research (CEBR) estimates businesses across Australia and New Zealand overspent on technology infrastructure by around AU$3.9 billion last year with those overspends forcing companies to scale back AI rollouts, freeze hiring and delay product development.

“Between 94 and 97 percent said opaque, unpredictable or difficult-to-forecast provider charges contributed to the overrun.”

The research, which surveyed 690 IT and infrastructure decision makers across eight countries, found 78 percent of Australian and New Zealand organisations experienced technology infrastructure costs that exceeded expectations in the previous 12 months. That was the second-highest rate recorded in the study, second only to France’s 79 percent, and above the global average of 71 percent.

The study filtered out overspend caused by an organisation’s own planned business growth, intentional usage spikes or internal budget miscalculations, instead isolating costs triggered by vendor pricing structures that organisations can’t reliably predict in advance.

For A/NZ the estimated annual provider-attributable infrastructure overspend was $2.3 billion for Australia and $0.5 billion for New Zealand.

The report, The Economic Cost of Unpredictable Technology Infrastructure, highlights a growing tension between technology consumption and financial forecasting. As organisations increasingly rely on pay-as-you-go cloud services, AI platforms and usage-based pricing models, costs can fluctuate according to activity levels, data movement, API requests and software consumption.

Among businesses that exceeded budget, between 94 and 97 percent said opaque, unpredictable or difficult-to-forecast provider charges contributed to the overrun. Respondents estimated those charges accounted for roughly one-third of unexpected infrastructure costs on average.

Faster-than-expected usage was prominent in storage, data and databases, compute and AI over spending, but fees such as data transfer, egress, retrieval, access and API charges were persistent pain points, with 43 percent of those experiencing storage overspend citing them. In AI environments, nearly half pointed to higher than expected AI platform, model or tool licensing costs.

Egress fees, interoperability barriers and licensing practices have been under scrutiny by regulators around the world in recent years. The UK’s Competition and Markets Authority 2025 cloud market investigation found Amazon and Microsoft held significant market power and limited customer choice through data egress fees, barriers to interoperability and Microsoft’s software licensing practices. Earlier this year the two vendors committed to reducing egress fees and improving interoperability.

The EU Data Act also specifically targets cloud switching barriers with caps for switching-related egress fees, mandated interoperability and open interfaces, while the US Federal Trade Commission has also held formal enquiries investigating the same issues.

The ACCC also highlighted competition risks in its Digital Platform Services Inquiry in 2025 saying regulatory reform was needed to improve competition.

Globally, the CEBR report shows AI and machine learning accounted for 37 percent, or US$23.4 billion, of the total US$63.4 billion overspend across the eight markets. Integration, deployment and model operations expenses were the second biggest cause of AI overspend, at 45 percent, behind licensing costs (49 percent) and ahead of token and usage charges (38 percent).

The opportunity cost

For Australian and New Zealand businesses already navigating tight economic margins and competition for digital talent, absorbing billions in unbudgeted tech spend carries a steep opportunity cost.

Owen Good, head of economics advisory at CEBR says the economic cost of unpredictable pricing is not simply that companies pay more.

“Our research uncovered the hidden cost of innovation for some of the world’s largest economies,” he says. “This means hiring and investment in new technologies are postponed or reduced as budgets are redirected to cover unexpected charges. The technology generating the most unplanned costs is also among the investments businesses are cutting to pay for them.”

Across the global sample, 84 percent of organisations reported having to make at least one cost-driven sacrifice over the past year to cover unexpected infrastructure bills. Nearly a third cut back on new software and technology investments (29 percent), staff training (28 percent) and hiring in critical tech, data and AI roles (27 percent).

The report says avoiding provider attributable infrastructure overspend could unlock $0.8 billion in additional capital-forming investment – or expenditure on R&D, AI initiatives and new technology that adds to an organisation’s productive capacity – for Australia and $0.2 billion for New Zealand. A further $0.3 billion in Australia and $0.1 billion in New Zealand could be redirected towards training and workforce development, while the funds allocated to hiring could support an additional 3,199 FTEs in Australia and 1,209 in New Zealand.

Commenting on the regional findings, Wasabi Technologies ANZ country manager Craig Stockdale notes that local organisations are increasingly penalised simply for accessing, moving, or scaling their own data, trapping IT budgets in operational maintenance rather than forward-looking innovation.

“Australian and New Zealand businesses are being taxed twice for innovation, once to build it, and again just to access or move their own data,” he says. “Businesses here tell us moving data between platforms or providers is a constant headache, and it’s fundamentally a pricing and architecture problem. Recovering that A$3.9 billion isn’t just about cost savings, it could be billions in AI and R&D investment and thousands of jobs this region is currently leaving on the table.”

More news in Artificial Intelligence
Latest news
iStart – Technology in Business

Helping business find and buy better software

About iStart Subscribe Advertise with iStart Contact iStart Privacy policy
NZ: 0800 928 268 · +64 21 796 723
AU: 1800 462 388 · +61 401 189 448
© iStart Legal · AU / NZ