State of Tokenomics: 472 company responses on the economics of AI spending

2026-09-23 · Publisher : Tokenomics Foundation · Domain : Cloud & DevOps

The Tokenomics Foundation publishes release candidate 1.0 of State of Tokenomics: three in four enterprises cannot confidently prove the outcomes of their AI spending to their CFO.

### Background On 23 September 2026, the Tokenomics Foundation publishes release candidate 1.0 of **State of Tokenomics**, a survey of 472 responses across 11 industries, nearly 500 companies representing USD 4.6 trillion of combined revenue. The study documents how enterprises measure, own, route and govern AI economics. The FinOps Foundation relays the topic: its 25 August 2026 article places "tokenomics" and FinOps for AI as two complementary disciplines projecting the same work along different axes. ### Key facts - Proof of value: three in four enterprises cannot confidently prove AI outcomes to their CFO; 39% are not confident they can link AI spend to a measurable outcome. - Ownership: 88% have defined an owner for AI economics; 35% name the technology function and 26% shared ownership. - Model routing: 86% are evaluating or using a model router; those using one are 4 times more likely to show CFO value. - Ask of providers: 23% ask for more granular spend data, only 4% for lower prices. - Pricing: 52% have already changed or are considering changing their pricing because of AI. ### Why it matters FinOps no longer covers only instances and storage: the AI bill, measured in tokens, becomes a governance topic. For IT departments and services firms, the survey points to workstreams to open: attributing AI costs to products, model routing, granular spend data from providers. The sample is self-reported and drawn from the foundation's community; this is a release candidate, not a final version. ### Takeaway The AI bill becomes a FinOps subject in its own right, with a central stake: proving value.

Tags : FinOps, IA, Cloud, Données