Apple cuts daily AI spending limits and budgets for engineering teams
Apple is cutting daily AI spending limits for engineering teams from $300 to $200 and plans further reductions to optimize internal costs.

According to Bloomberg, Apple is cutting the daily budget that its engineering divisions can spend on artificial intelligence technologies. This marks a significant shift for a company that has been actively integrating AI into its core engineering processes over the past year.
- In October, Apple lowered the daily AI spending limit for engineering teams from $300 to $200.
- Further reductions are planned, potentially dropping the limit to $100 per day for certain groups.
- Non-engineering departments may face an even stricter limit of $25 per day.
- Developers are adjusting their work schedules to utilize AI tokens when they are available.
The Context Behind AI Spending Cuts
Earlier this year, the tech industry experienced a massive AI boom, with companies urging employees to use AI tools as frequently as possible under the guise of boosting efficiency. However, the lack of spending caps led to astronomical costs. For instance, an unnamed US company spent $500 million in a single month on Anthropic's Claude tokens because no employee limits were established, while Meta employees burned through 73.7 trillion tokens in a month.
Impact on Developer Workflows
Read alsoAnthropic bans Claude resale and proxy accessThe newly imposed restrictions are already altering the daily routines of software engineers. Developers have reported changing their work schedules to maximize utility from AI tools, specifically starting their workday during hours when AI tokens are readily available. This highlights how deeply reliant modern technical staff have become on generative AI for coding and problem-solving.
What This Means for the Industry
Apple's decision to rein in internal AI budgets highlights a broader transition in the tech sector from unchecked experimentation to strict financial discipline and cost optimization. As companies like Meta also implement strict spending caps, it becomes evident that the high operational costs of large language models require careful monetization and strict internal monitoring.
Source: Все новости подряд



