AI Cloud Costs Expose FinOps Gap As Only 13% Fully Allocate Spend

IBM Apptio’s latest technology investment findings put a sharper number on the cloud and AI cost problem: only 13 percent of FinOps respondents can trace AI and machine-learning cloud bills to business owners and pair that allocation with optimisation insight.

In an interview with iTNews Asia, Pete Wilson, IBM Apptio Business APAC vice president and general manager, framed that gap as a business value problem rather than a narrow cost-control issue.

Technology leaders can usually see money moving through infrastructure, cloud platforms, software and AI initiatives, but the harder task is explaining which spending supports operational improvement, service delivery or strategic objectives.

The pressure is rising as cloud and AI move from pilots into broader deployment.

IBM Apptio’s recent 2026 global Technology Investment Management report found that about half of respondents had low confidence in cloud spend forecasting, while AI investments had driven significant cost increases and overruns over the past 12 months.

That uncertainty changes the work of finance and technology teams.

When forecasts miss, leaders spend more time reconciling variance and less time directing investment.

Wilson pointed to public cloud, AI vendors and labour as the cost pools that need closer tracking, with weekly or even daily cloud-spend monitoring used to catch early signs of growth from expanding AI usage.

Legacy planning systems remain part of the constraint.

ERP platforms remain the primary budget system for nearly half of organisations, compared with about a third using purpose-built IT financial management platforms and 7 percent relying on spreadsheets or other manual processes.

The source described those older models as built for more stable, asset-based technology environments, not the faster planning cycles created by public cloud and AI.

The allocation problem becomes more visible when AI and cloud services spread across teams.

IBM Apptio’s report found that nine out of 10 leaders ranked AI as a priority investment area, with cloud services among the most closely watched spending categories; it put full cloud-cost chargeback at only about one in 10 organisations.

Wilson linked that allocation gap to usage spread across shared services, hybrid environments and business units.

IBM Apptio’s regional view also suggested that tooling alone is not the barrier.

In the region, FinOps work is still often anchored in providers’ native cost consoles, creating blind spots across hybrid and multi-cloud estates.

Leaders may know overall costs are rising, but fragmented data makes it harder to tie a workload, application or business decision to the increase.

Confidence can also run ahead of capability.

Nearly three in five IT financial management professionals believed their forecasts were highly accurate, even though only about a third used purpose-built tools to manage and validate technology spend.

In cloud, hybrid and AI environments, disconnected systems and manual processes can leave teams with partial visibility rather than a single view of consumption, unit costs and business ownership.

The budget source matters as much as the forecast.

Existing budgets, not fresh capital, are the main source for two-thirds of polled initiatives, increasing pressure to shift money continuously instead of relying on annual plans.

That dynamic makes the “run” versus “growth” trade-off more immediate: savings from operating costs may be needed to fund new AI, cybersecurity or cloud-modernisation work.

Wilson’s practical threshold was monthly, digitally delivered charging that lets business consumers see not just what they owe, but the units consumed and price per unit.

Wilson described CIOs and CFOs without that transparency as making “best guesses” about where to invest.

The operating condition for scaling cloud and AI is therefore less about adopting another tool and more about unifying cost, usage and delivery data before spending growth outruns decision confidence.