Microsoft Cancels Most Licenses for AI-Powered Coding Tool
Due to soaring expenses, Microsoft has decided to cancel the majority of licenses for Anthropic's Claude Code, an AI-driven programming assistant. This move is part of a broader strategy that includes shifting engineers to GitHub Copilot CLI instead. Despite the license cancellations, Microsoft remains committed to investing in Anthropic, with plans to inject up to $5 billion. In return, Anthropic has pledged to spend $30 billion on Azure cloud services.
While Microsoft scales back its use of Claude Code, Uber has reported exhausting its entire annual budget for AI tools in just four months. Uber’s Chief Technology Officer, Pravin Neppalli Naga, commented:
“I’m back to zero because the budget I planned is already used up.” – Pravin Neppalli Naga
This situation highlights that roughly 70% of Uber’s code is now written with AI assistance, underscoring the growing role of artificial intelligence in software development. Individual engineers at Uber spend between $500 and $2,000 per month on tokens to access AI tools. The token-based payment model is raising concerns, as even when the price per token drops, total costs continue to climb because usage increases. This observation is echoed by Fortune.
Rising AI Costs in the Corporate Sector
A similar trend of escalating AI expenses is emerging across the corporate world. Gartner has warned about rising enterprise AI costs through 2030, while Goldman Sachs projects a 24-fold increase in token consumption by that year, with monthly usage expected to reach up to 120 quadrillion tokens. Meanwhile, studies indicate that over 80% of AI projects in business fail to deliver expected results, highlighting the risks tied to investments in new technologies.
The cancellation of Claude Code licenses may signal the growing financial hurdles tech companies face with AI adoption. At the same time, Uber’s experience shows that even increasing reliance on AI can lead to budget overruns, forcing companies to rethink how they fund such technologies. In light of these developments, the need for strategic planning and risk assessment in AI implementation is becoming increasingly clear for the corporate sector.