Dive Brief:
- AI sprawl is undermining IT asset management and cost control measures, according to a Flexera study published Wednesday. The software company surveyed more than 500 IT professionals with asset management responsibilities in early 2026.
- More than two-thirds of respondents lacked accurate visibility into AI software usage and just 36% said they had a complete picture of their organization’s IT assets. Roughly 3 in 5 acknowledged AI overspend had increased year over year, with wasted spend on SaaS growing 10 percentage points in the last year.
- “AI is changing the economics of IT faster than most organizations can adapt,” Becky Trevino, chief product officer at Flexera, said in a release accompanying the report. “What we’re seeing is a familiar pattern of rapid adoption followed by a scramble for visibility and control as spend surges.”
Dive Insight:
The race to ease workflows and wring efficiency gains through rapid deployment of generative AI has created a tangle of governance and security risks. But unrestrained spending on the technology took center stage last month amid reports that Uber tore through its entire 2026 AI budget in just four months.
The Uber incident was one example of tokenmaxxing, an AI usage practice described in a Forrester blog post as “what you get when you incentivize adoption without governing value.”
Budget overruns are a predictable outcome of the rushed deployment of an emerging technology, according to Trevino.
“Like most things in the corporate world, it became a top-down initiative,” Trevino told Channel Dive. “It went from giving everyone the latest Microsoft subscription to rewarding whoever was using the most tokens. And as we started to use AI irresponsibly, the bills started shooting up.”
Tracking and adopting AI was the top challenge for ITAM leaders, according to the Flexera survey. More than half of large organizations polled bundled AI spending into software budgets, and 7% said they weren’t currently monitoring the technology’s usage costs separately.
Most businesses have more on the line than a company with Uber’s profile, Trevino said.
“If their technology budget is outgrowing the pace at which the revenue is coming in, that's a dangerous place to be,” she said. “If you're a tech startup, that's normal. But if you're a manufacturing organization out of Ohio, that's not okay … a lot of people are asking, ‘Where's the 30% productivity I was promised from all the Microsoft Research? I'm not seeing it.”
ITAM frustration has spilled over into the channel as organizations turn to partners for AI deployment, governance and budgeting guidance. Three-quarters of the organizations surveyed by Flexera contracted some or all of IT asset management responsibilities to managed service providers.
Forrester expects the need for managed AI services to expand as agentic automation spreads. Organizations will lean on MSPs to operate and optimize AI agents, the analyst firm said.
The FinOps Foundation, which added AI to its cost management agenda last year, took on token-based pricing through its parent organization, the Linux Foundation, earlier this month. The nonprofits launched the Tokenomic Foundation to urge vendors to rationalize AI billing practices and will hold an inaugural Tokenomicon convention next June in San Diego.
“Most organizations still lack the maturity and frameworks to effectively measure cost versus business impact,” Gartner Senior Principal Analyst Nitish Tyagi said in a Wednesday report. “Software engineering leaders are increasingly concerned as token-driven AI spend becomes harder to justify, with budgets often being depleted earlier than expected.”
Gartner expects AI coding agent costs to surpass the average developer’s salary by 2028, largely due to rising token costs.
