Weak data practices waste AUD $154bn in global AI spend

Hitachi Vantara has published a new survey on data infrastructure readiness that links weak data management practices with an estimated AUD$154 billion in global AI investment waste each year.

The State of Data Infrastructure Report draws on responses from more than 1,200 C-level executives and IT leaders across 15 countries.

The survey includes 80 respondents in Australia.

In Australia, 56% of organisations sit in lower maturity stages for data management, according to the survey. The report classifies this group as operating with weaker data infrastructure practices. It says AI work has exposed gaps in data management, governance and security.

Data complexity also emerged as a central theme in the local results. The report found that 78% of Australian organisations said complexity in their data infrastructure environments is growing rapidly or too quickly to manage. The report attributed complexity growth to increases in data volumes, platforms and AI activity.

Hitachi Vantara also reported an expected increase in AI spending. It said businesses anticipate AI investment will grow by 69% over the next two years. The report linked this trajectory with rising pressure on governance and security.

Visibility and control

The report described an operational challenge that follows complexity. It said organisations struggle to maintain visibility, control and accountability across systems as environments expand.

Several survey results focused on operational readiness and risk.

Only 49% of Australian respondents reported having predictive or automated infrastructure operations. The report said this limits an organisation's ability to manage complexity.

Security and resilience concerns also featured in the responses. Some 58% of Australian business and IT leaders said complexity in their data makes identifying a data breach more difficult. A further 54% said a critical data loss would be catastrophic.

The report also found that 50% of Australian respondents said their systems are complex enough that executives would lose sleep if they understood the risks.

One of the authors linked the findings with how organisations run data governance. "AI is raising the bar for how organisations govern and manage their data," said Octavian Tanase, Chief Product Officer, Hitachi Vantara.

Tanase said business leaders now focus on governance in parallel with other operational factors. "As AI becomes more embedded in business operations, leaders are realising that governance, visibility and control matter just as much as performance," said Tanase.

He also contrasted different levels of readiness across organisations. "As a result, we're seeing that organisations that have invested in automation and optimised data infrastructure are moving faster with confidence, while others are seeing complexity widen the gap between those that can manage it effectively and those that cannot," said Tanase.

Readiness divide

The report found near-universal AI engagement across the global sample. It said 95% of organisations are using, piloting or exploring AI.

It drew a distinction between adoption and outcomes at scale. The report said readiness to scale AI and realise value varies widely across organisations. It described a divide between organisations with strong data management foundations and those struggling to keep pace.

The report categorised 44% of organisations as data-mature. It defined this group as having managed or optimised data practices. It categorised 56% as data laggards. It said these organisations operate in defined, emerging or fragmented stages of data management.

In Australia, the report linked outcomes to data quality and compute constraints. It said 71% of organisations had success using AI. It also said 36% cited the use of incorrect data as a top reason for unsuccessful AI projects. Another 28% cited a lack of processing power as a leading factor behind unsuccessful work.

Survey respondents also identified what contributed to success. The report said data quality was the most commonly cited driver of successful AI initiatives. Some 42% of Australian organisations attributed successful AI projects to the use of high-quality data.

The report also suggested AI has moved into core operations for a significant share of organisations. It said 35% of Australian organisations now consider AI a critical part of their function.

Automation and resilience

The report compared data-mature organisations with those it categorised as less mature. It said the biggest differences appeared in automation, infrastructure modernisation and operational discipline.

In Australia, the survey found 81% of organisations with stronger data foundations operate predictive or automated infrastructure. The figure drops to 25% among organisations with less mature data practices.

The report also measured design and resilience approaches. It said 73% of data-mature organisations in Australia report having sustainable design and built-in resilience. It said 18% of less mature organisations report the same.

Another theme in the report centred on leadership engagement. It said leadership alignment remains a distinguishing feature between data-mature organisations and others.

The report said 95% of Australian organisations say they need outside help with data infrastructure. It also said many organisations struggle to turn that need into coordinated action.

"As AI becomes central to how every business operates, leadership has to treat data foundations as a strategic requirement, not just a technical concern," said Sheila Rohra, Chief Executive Officer, Hitachi Vantara.

Rohra also linked AI performance with governance and resilience. "This report makes clear that AI succeeds when the data behind it is trusted, well-governed and resilient," said Rohra.

She described Hitachi Vantara's focus areas for customers. "Our role is to help organisations simplify the management of their environments, strengthen governance and make sure their data strategy can support long-term growth," said Rohra.