Archive — US — techcrunch
The token bill comes due: Inside the industry scramble to manage AI’s runawaycosts
As AI costs soar, companies like Uber and Microsoft are grappling with massive overages in token spending, prompting a search for cost-control solutions. The new Tokenomics Foundation aims to standardize token economics, while startups and established vendors develop tools to track and optimize AI expenses. With global token usage projected to burgeon by 2030, urgent measures are being sought to manage these escalating costs.
Key points
- Uber exhausted its 2026 AI coding budget by April 2025.
- Microsoft revoked Claude Code licenses due to rising costs.
- Companies that initially embraced AI are now reeling from expenses and seeking visibility and controls.
- Increased demand for AI has led to heightened token consumption despite falling per-token prices.
- The Tokenomics Foundation was launched to bring rigor to AI token management akin to financial operations for cloud spending.
- Some companies report overages reaching 3x their budgets early in the fiscal year.
- Productivity gains from AI use are countered by exponential spending; firms find it hard to measure ROI.
- The market is evolving with companies creating tools to manage AI spending, reflecting a need for a common language in token usage.
Why it matters
- The conversation has shifted from innovation in AI to managing expenditures.
- Historical parallels to telecom and cloud expenses highlight the typical challenges of new tech adoption.
- There exists a trend towards greater scrutiny and optimization in costs associated with emerging technologies, similar to cloud management practices.
With AI token costs outpacing initial budgets, companies face a critical need for tracking and optimizing expenses. The formation of the Tokenomics Foundation signals a pivotal step towards establishing standardized practices that could bring order to a rapidly evolving landscape.