📰 Key Takeaways
Apple is reportedly cutting the daily AI budget available to its internal engineering teams. Bloomberg reported on Wednesday that in October Apple lowered the daily AI spending cap for engineering teams from $300 to $200, with plans to cut it further, and some teams will see their allowance fall to $100 a day. Sources also said that for units outside engineering, Apple is evaluating a daily cap as low as $25. That’s a big shift for a company that has long worked to weave AI into its engineering workflows. The report notes that earlier this year there was a “tokenmaxxing” trend, where many companies encouraged employees to use as much AI as possible in the name of vaguely defined efficiency gains. But the craze didn’t last long. As the bills for token costs started coming due, companies began to see capping token usage as the financially sounder move. Doubts linger too: there still seems to be a gap between what it costs to build and run AI and the productivity gains the technology claims to deliver, which are often hard to measure clearly. TechCrunch has asked Apple for comment and hasn’t heard back yet.
💬 JudyAI Lab Take
Apple is reportedly cutting the daily AI budget for its internal engineering teams, and it’s worth paying attention to. It shows the “use as much AI as you can” craze getting pulled back to reality by the bill, with even a company that has long integrated AI into its engineering workflows starting to hit the brakes.
For AI builders, this case reflects a shift in design thinking: tokens are no longer an unlimited resource, they’re a cost line you have to plan for. According to the original report, Apple has already lowered the daily cap for engineering teams from $300 to $200, some teams will go as low as $100, and non-engineering units could be squeezed to $25. At the same time, doubts remain about whether there’s a gap between the cost of building and running AI and the productivity gains that are hard to measure clearly. We think this means future AI workflows can’t just ask “can we do it?” They also have to ask “is each call worth it?” A design that gets the job done with fewer tokens will hold up better than one that just stacks up more model calls.
My suggestion: this week, take stock of your own projects’ AI usage, find the three steps that burn the most tokens, and write down the concrete output each one delivers.
📅 Source Info
- Published: 2026-11-05T01:26
- Original article: https://techcrunch.com/2026/11/04/apple-is-reportedly-cutting-budgets-for-its-internal-ai-usage/