Apple has tightened spending on internal AI tools, reducing daily budgets for some engineering teams from $300 to $200 and encouraging the use of cheaper models or on-premises solutions.Author and source: AIBase
According to multiple media outlets citing insiders, Apple has begun tightening the cost of its internal AI tools. The most immediate changes affect engineers: in October, the daily AI budget for some engineering teams was reduced from $300 to $200, and the company is exploring further cuts, with some teams potentially facing a limit of just $100 per day in the future. By comparison, many non-engineering teams have already been operating under a $100 cap, and internal discussions have even considered reducing some teams’ limits to $25 per day. Employees whose roles genuinely require higher quotas can still request additional allowances from executives.
What was cut was waste, not strategy.
These budgets primarily cover token consumption for generative models like Claude, ChatGPT, and Gemini—engineers, who heavily rely on AI for programming, have always been allocated the highest tiers. Over the past year, the Silicon Valley trend of “Token Maxxing” has driven usage sharply higher, with some even competing to see who uses the most. To curb this trend, Apple is encouraging teams to switch to more cost-effective models or prioritize AI solutions running locally on devices. As tokens are the standard billing unit for mainstream AI services, higher usage means larger bills—this is precisely where Apple is tightening the valve. However, it’s worth noting that the company’s overall R&D spending still increased significantly year-over-year in fiscal year 2026, driven largely by AI and infrastructure, indicating that the strategy has not slowed—only become more disciplined in usage.
The industry has reached its accounting moment.
This shift from “spending freely” to “accounting for costs” is not unique to Apple; over the past year, the industry experienced massive expansion, and now companies are asking how much productivity these expenditures have delivered. Internal attitudes are divided: some believe the $300 limit was already generous and rarely reached, making the new rules largely inconsequential; others see it as a precursor to broader cost-cutting measures, especially as certain business units have brought in external consultants to evaluate efficiency—fueling quiet speculation about restructuring, though there is currently no evidence that Apple has finalized any layoff plans. After the early surge, companies are shifting their focus from “how much they’re using” to “whether it’s worth it”—controlling operational costs has become a shared new challenge across the industry.

