BlockBeats News, May 29th, as enterprises scale up AI deployment, more and more companies are beginning to question whether the high AI spending truly brings a return on investment. Reports indicate that Microsoft has canceled most of its Claude Code licenses, partly due to the high cost; Uber's COO has also publicly stated that AI spending is becoming increasingly "difficult to rationalize."
Of more concern, an AI consultant revealed to Axios that one of their clients, due to not restricting employee use of Claude, ended up spending as much as $500 million in a single month.
Industry insiders point out that many companies had previously fallen into the so-called "tokenmaxxing" trap—meaning they were aggressively consuming AI tokens, striving for full organizational AI transformation, but the actual return on investment was not ideal.
Micro1 CEO Ali Ansari stated that in the current truly mature large-scale business scenarios for AI, "it's mainly about programming," but many companies blindly apply AI across the entire organization, leading to a sharp increase in IT costs.
The report also mentioned that some enterprise employees would even use high-cost AI models to query simple questions like the weather, and enterprise-level AI services are not truly "unlimited," as a large number of simple requests will also rapidly drive up token costs.
Meanwhile, more and more companies are starting to offset AI spending pressure through layoffs. CloudBees CEO Anuj Kapur stated that the real reason for some companies' layoffs may simply be that "this is the only way to cover the AI bill."
---------------------------------
Click the original text link below to join the BlockBeats · Feishu AI News Channel, monitoring global AI hotspots and news 24/7.
