Apple is reportedly cutting the daily amount its engineering teams can spend on AI, according to Bloomberg, as cited in the report. The company lowered the cap from $300 to $200 a day in October, and some allowances may fall to $100 a day. Outside engineering, Apple is reportedly weighing a daily limit as low as $25. For a company that has worked to build the technology into its engineering processes, these are significant reversals.
The numbers matter less than what they reveal. The report describes a pattern in which companies pushed staff to use AI as much as possible in the name of vaguely defined efficiency gains. That frenzy, the report says, lasted only until the bills for tokens came due. When spending becomes visible, companies start rationing. Apple's move looks less like a retreat from AI than a belated acknowledgment that usage was never tied to a clear return.
The report also points to a broader concern: an apparent gap between what AI costs to build and run and the productivity gains it is supposed to generate. Those gains are often described in unclear terms, and the report does not offer a measured figure for them. Until a company can show what its token spending produces, budget caps are a reasonable discipline, not a sign of weakness.
Apple should go further than cutting numbers. Each team that consumes AI budget should be asked to name the task, the expected output, and how success will be measured. Spending that cannot be tied to those answers should be the first to go. Caps alone do not tell staff what good use looks like.
The report notes that TechCrunch reached out to Apple for comment, and it does not say whether the company has responded. Any judgment about the cuts should wait for Apple's explanation. Still, the direction is clear: the industry's appetite for AI spending is being tested against its bills, and the test will be decisive.