Microsoft Pulled Back on Claude, Uber Burned Its 2026 AI Budget by April: The Inference Cost Wall Is Here
Reports that Microsoft cut its use of Anthropic's Claude over exploding costs — and that Uber blew through its entire 2026 AI budget in four months — signal that enterprise AI economics are hitting a hard ceiling far sooner than anyone planned.
The AI cost crisis that infrastructure teams have been whispering about for months is no longer theoretical. Multiple reports circulating this weekend paint a stark picture: Microsoft reportedly scaled back its use of Anthropic's Claude models after inference bills exploded beyond internal projections, as @0xsadikbaba documented in a widely shared thread. Uber, meanwhile, allegedly torched its full 2026 AI budget by April — barely a third of the way through the fiscal year.
The details remain patchy, and neither company has issued formal statements confirming the specifics. But the directionality is consistent across multiple observers. As @eager28 put it bluntly: "We always wondered where the wall was. Now we found it." The wall, it turns out, is not a technical constraint or a talent bottleneck — it's the spreadsheet.
Get our free daily newsletter
Get this article free — plus the lead story every day — delivered to your inbox.
Want every article and the full archive? Upgrade anytime.
No spam. Unsubscribe anytime.