Enterprise AI spending faces its first real ROI audit season
After two years of open-ended budget, finance teams are asking software and AI leads to show the math.

Finance teams at a number of large enterprises are running their first formal return-on-investment reviews of AI tooling purchased over the past two budget cycles, after a period in which AI spend was largely exempted from the scrutiny applied to other software categories.
The reviews are turning up a wide spread of outcomes: tools embedded in high-volume, well-defined workflows show measurable time savings, while broader "AI assistant" rollouts with vaguer use cases are proving harder to attribute value to.
Some of the harshest findings have come from license utilization data rather than productivity metrics — several companies discovered seat utilization well below 50% on tools purchased for company-wide rollout.
The net effect heading into the next budget cycle is consolidation rather than retreat: fewer, more targeted AI tools tied to specific workflows, replacing the broader experimental footprint built up over the last two years.

