Apple set to slash employee AI token spending as new CEO John Ternus looks to cut costs

Apple prepares to reduce artificial intelligence spending per employee, according to Bloomberg. The tech giant aims to curb unpredictable costs that have been steadily rising. Previously, Apple capped AI tool expenditures at $300 per engineer per day, but this figure is expected to drop to as low as $100. Non-engineers could face even stricter limits, with a proposed cut from $100 to $25 per day.

The reductions are reportedly being implemented across the company, though the specifics may vary based on team and project requirements. Apple has not yet commented on the news. To mitigate the impact, employees may be encouraged to use cheaper AI models, such as those offered by ChatGPT, Claude, and Gemini, instead of the more expensive alternatives.

Additionally, engineers might be asked to run workloads locally on the company's own hardware. This shift echoes similar cost-cutting measures taken by other major tech firms, such as UK digital bank Revolut, which recently switched from Claude to open-weight models hosted on its own infrastructure. While the proposed budget cuts may appear substantial at first glance, the reality could be less severe for some workers.

Apple's AI expenditure has already seen a 32% increase in R&D costs from FY 2025 to FY 2026, and total operating expenses rose by 22%. Despite this, the company's quarterly revenue only grew by 12% year-over-year.