Dive Brief:
- In a sweeping effort to establish AI cost standards, the Linux Foundation launched the vendor-neutral Tokenomics Foundation on Tuesday. Among the group’s 30 founding members are several major channel players, including Accenture, IBM, SAP, DoiT, WWT and SHI.
- The move comes as companies grapple with rising token costs, AI overspend and increasing demand for measurable business value. The nonprofit foundation will define value metrics for AI ROI, create frameworks tying token spending to business outcomes and deliver standard methods for measuring AI profit.
- “Most organizations that can allocate cloud spend to the dollar cannot tell you what a single AI feature costs to run, so they cannot answer the question their boards are asking: ‘what is this AI investment returning?’ Amit Kinha, Field CTO at DoiT and Tokenomics Foundation governing board member, said in an email. “The Foundation's job is to give the industry a shared, vendor-neutral way to measure, attribute and value AI consumption so that question becomes answerable.”
Dive Insight:
AI is upending what companies know about managing and monetizing their investments. Organizations are overinflating tech budgets while losing track of token spending. Many are continuing to invest in AI despite minimal measurable returns on investment. By 2030, token consumption is projected to increase 24-fold, according to Goldman Sachs research.
The Tokenomics Foundation, which is part of the Linux Foundation, represents an industry-wide call for the rapid establishment of best practices for AI economics.
“Every framework we have for measuring technology spend assumes that whoever provisions a resource is whoever consumes it,” Kinha said. “AI broke that assumption. Spend now follows tokens, model calls and GPU cycles that scatter across teams, features, agents and customers, and no existing standard was built to measure any of it. This is a new economic layer, and it needs its own discipline.”
Some partners have taken it upon themselves to rein in spending. Cloud-based IT marketplace provider Pax8 announced it would build token-tracking capabilities into its operations platform last month. The company aims to help managed service providers simplify billing for AI tools based on token consumption.
With its tokenomics initiative, the Linux Foundation wants to follow a roadmap drawn by the FinOps Foundation, which rallied industry support for cloud billing standards and cost controls. Tim Brooks, area VP of AI development and business advisors at WWT, told Channel Dive the Tokenomics Foundation will help organizations make better architecture and operating decisions while optimizing AI consumption and ROI.
“As enterprises move AI from experimentation to production, they need a FinOps model built for AI’s unique unit economics — including tokens, GPUs and inference — to govern spend and demonstrate business value,” Brooks said in an email.
Helping to define the economics of AI is a prime channel opportunity, according to Shane Cronin, head of spend optimization services at SHI.
“Where most of the opportunity sits for channel providers today is being able to address the client challenge of managing the economics of AI,” Cronin said. “But there certainly won’t be space for everyone to have a solution. That’s why we engaged early — we want to be part of not just defining what tokenomics is, but also building all of the frameworks and influencing the industry as it transforms.”
