Tokenomics Foundation launches to curb runaway AI token costs

Across the tech world, companies are starting to balk at the price of artificial intelligence. In response to these expenses, the Linux Foundation has unveiled plans to launch the Tokenomics Foundation. The new standards body aims to instill cost discipline around tokenomics—the economics and cost management of AI token usage—mirroring how the cloud financial management practice known as FinOps brought order to cloud computing spend.

The financial strain has hit enterprises. In early 2025, companies gorged themselves on all-you-can-eat subscriptions, but they are now facing bills. For instance, Uber blew through its entire 2026 AI coding budget by April. Another company reportedly faced a $500 million Claude bill after failing to set usage limits. At Priceline, a routine contract renewal for the coding assistant Cursor came back 4-5x more expensive. This increase is driven by the rise of AI tools capable of autonomous action, which have pushed per-developer token consumption up by about 18.6x over a nine-month period.

This inflation of costs has shifted corporate priorities. Alexander Embiricos, OpenAI’s head of enterprise, noted at an event in New York City that customer conversations have shifted from model capabilities to cost visibility, auditability, and token controls. Chris Reed, senior director of IT finance at Priceline, described the sudden dependency on these models: “It’s like the crack-cocaine epidemic.” To manage this, Priceline has begun placing token limits on certain groups, while other firms are grappling with individual usage, such as a single engineer spending $40,000 on tokens in one month.

A market of monitoring tools is forming to help companies track spending and prove return on investment. Engineering management platforms like Jellyfish and engineering operations platforms like Faros AI are analyzing developer habits. A study of 20,000 developers by Faros AI found that while output rose, so did bugs and rewrites. Meanwhile, Jellyfish found that engineers using the most tokens were twice as productive but consumed 10x more tokens to get there. Despite these efficiency questions, the demand for tokens shows no signs of slowing down; Goldman Sachs projects that global token usage will multiply by 24 times by 2030.

Why it matters

As AI adoption shifts from experimentation to production, the lack of standardized cost management is creating an existential budget crisis for enterprises, necessitating new industry-wide frameworks.