AI sprawl, token consumption ratchet up tech overspending

· Source: Information and Enterprise Technology News | CIO Dive - Www.ciodive.com · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cloud Computing & IT Infrastructure, Emerging Technologies & Innovation · Depth: Intermediate, short

Summary

A Flexera study, published June 29, 2026, reveals that AI sprawl and increasing token consumption are driving significant tech overspending, with nearly two-thirds of organizations lacking adequate IT asset visibility to control these costs. The survey of over 500 IT professionals in early 2026 found only 31% have visibility into AI software usage, and just 36% possess a complete picture of their IT assets. Consequently, roughly three in five respondents reported increased AI overspend year over year, alongside a 10 percentage point rise in wasted SaaS spend. This trend, exemplified by Uber depleting its 2026 AI budget in four months due to "tokenmaxxing," highlights a rapid adoption pattern followed by a scramble for control. Many organizations are now contracting IT asset management responsibilities to managed service providers, with three-quarters of surveyed companies doing so. Furthermore, the FinOps Foundation and Linux Foundation launched the Tokenomic Foundation to address AI billing practices, while Gartner predicts AI coding agent costs will exceed average developer salaries by 2028.

Key takeaway

For Directors of AI/ML or IT Asset Management struggling with escalating AI expenditures, you must prioritize gaining granular visibility into AI software usage and token consumption. Uncontrolled AI sprawl can quickly deplete budgets, as seen with Uber, and erode promised productivity gains. Implement separate monitoring for AI costs and consider engaging managed service providers to establish robust governance frameworks and optimize spending before overruns become critical.

Key insights

Rapid, ungoverned AI adoption and token consumption are causing widespread IT overspending and visibility challenges.

Principles

In practice

Topics

Best for: CTO, VP of Engineering/Data, IT Professional, Director of AI/ML, Executive

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Information and Enterprise Technology News | CIO Dive - Www.ciodive.com.