355
Audio & Video Production344
Automation & Workflow224
Software Development250
Marketing & Growth192
AI Infrastructure & MLOps174
Writing & Content Creation203
Data & Analytics140
Design & Creative169
Customer Support131
Photography & Imaging156
Sales & Outreach125
Voice & Speech135
Operations & Admin87
Education & Learning131
Businesses say AI tool bills are climbing fast, so they are adding limits and looking for new ways to track and control “tokens,” the units AI uses.
In short: Companies are cutting back and adding controls as the cost of using AI tools, measured in “tokens,” is rising faster than many budgets.
Many companies are finding that everyday use of AI is getting expensive. TechCrunch reports that Uber used up its entire 2026 budget for AI coding tools by April. Microsoft also revoked some developers’ access to Claude Code months after allowing it.
The key issue is “tokens,” which are small chunks of text that AI systems read and write (like a meter counting every word you send and receive). Prices per token have generally fallen, but companies are using AI more often and for bigger tasks. Newer “agent” features, meaning AI that can take multi step actions on its own (like an assistant that keeps going after the first request), can run up token use quickly.
Some examples in the report show how hard it is to predict and control these costs. A Priceline employee said a Cursor contract renewal came back 4 to 5 times more expensive. One company reportedly faced a $500 million bill after forgetting to set usage limits.
Because of this, new tools and services are appearing to help track spending, compare vendors, and set guardrails. The Linux Foundation also announced plans for a new group called the Tokenomics Foundation. It aims to create shared definitions and standards for measuring token usage and billing, similar to how “FinOps” helped companies manage cloud computing bills.
Expect more companies to set stricter limits, and to push vendors for clearer reporting. Also watch whether the Tokenomics Foundation’s standards, planned to formally launch in July, make it easier to compare AI costs across providers and prevent surprise bills.
Source: TechCrunch AI