Only 31% of companies can see their AI software spend

Before a company can build a cost dashboard, it has to know what it's paying for. A June 2026 survey of 512 IT asset management professionals found only 31% have accurate visibility into their AI software spend — and 59% say wasted AI spend increased over the past year anyway.

The blind spot is upstream of the dashboard

Flexera's 2026 State of ITAM Report, published June 24, 2026 from a survey of 512 technology professionals worldwide, found that only 31% of organizations report accurate visibility into their AI software spend. Nearly half already "track" AI as part of their software spend inventory — but tracking isn't the same as accuracy. The report also found overall IT asset visibility dropped to 36%, a broader erosion, not one specific to AI.

That distinction matters more than it sounds. A dashboard built on top of an inventory that misclassifies what's an AI subscription and what's a general SaaS line will report numbers that look precise and are quietly wrong — the error happened before any monitoring tool ever ran.

Wasted spend keeps rising anyway

59% of respondents say wasted AI spend increased year over year, despite growing organizational awareness of the problem. The two figures fit together: visibility gaps compound rather than average out. If the underlying asset inventory doesn't correctly categorize a line item as AI spend, no amount of downstream cost monitoring catches the waste, because it was misfiled before monitoring ever had a chance to see it.

Chargeback breaks when the inventory is wrong

Organizations increasingly try to allocate AI cost back to the business unit that generated it — chargeback or showback, rather than absorbing everything into a central IT budget. In principle, API-based pricing makes this straightforward: consumption is billed per token, and per-team attribution is a matter of tagging requests correctly. In practice, that model only works if the underlying spend is correctly identified as AI in the first place. A line item folded into a general cloud or SaaS category can't be charged back accurately, so the accountability loop that's supposed to curb waste never actually closes.

Two different visibility problems, one blind spot

It's worth separating two layers that get conflated under "AI cost visibility." One is usage-level monitoring — which model, which request, which feature is spending, tracked in near real time. The other is asset-level visibility — whether the organization even has an accurate inventory of the AI software and API subscriptions it's paying for in the first place. The Flexera data points squarely at organizations still missing the second, more basic layer: you can't monitor usage accurately on top of an inventory that doesn't know what it's counting.

What accurate AI asset visibility requires

• A single inventory that correctly tags AI API and software line items instead of folding them into general cloud or SaaS categories.
• Reconciliation between what finance is billed and what's actually provisioned and active — the gap between the two is where wasted spend hides.
• Attribution granular enough to support chargeback, not just a total that tells finance the number without telling anyone which team owns it.

The bottom line

A cost dashboard is only as accurate as the inventory underneath it. The Flexera data suggests most organizations built the dashboard first and are still missing the inventory — which means the 69% without accurate AI spend visibility aren't necessarily lacking monitoring tools. They're lacking a correct starting list of what those tools are supposed to be watching.