Breaking

EY Executive Says AI Spending Is Facing a Harder ROI Test as Token Costs Rise

September 25, 2026 at 09:00 EDT

Sales & Business
AI Agents
Software Dev & Coding

Businesses are continuing to invest in AI, but the financial proof points remain uneven as usage-based pricing makes the cost of deployment harder to ignore. EY Global Consulting AI Leader Dan Diasio said many clients are still unable to point to clear AI-driven revenue gains or cost reductions, while EY survey data shows executives are now rethinking access, budgets and rollout plans around token consumption.

Sep. 25, 2026 · Enterprise AI

AI spending meets the ROI test

Companies are still funding AI, but token costs and usage-based pricing are forcing executives to prove where productivity becomes revenue, savings or operating leverage.

82%

Concerned about token usage and related costs

98%

Reconsidering AI approach because of token costs

64%

Monitoring token use with budget guardrails

1 in 10

Can show where AI ROI appears in income statements

The cost-control gap

The tallest columns show concern and reassessment. The shorter column shows how many organizations say they have active token monitoring and clear guardrails.

82%Concern

98%Reassessment

64%Guardrails

Why leaders keep going

Teams report productivity gains, faster internal software development and targeted gains in workflows such as coding, document analysis and customer operations.

Why CFOs are cautious

Productivity is not enough unless it reaches revenue, cost reduction or measurable profit impact, especially as AI bills become variable operating costs.

The new AI finance loop

Meter usageTrack tokens, API calls and agent runs.

→

Assign valueTie work to revenue, savings or quality.

→

Scale selectivelyExpand the workflows that prove return.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.