Apple cuts employee AI budgets as ChatGPT, Claude usage costs mount

Apple is reducing employee access to AI as part of a broader cost-cutting effort, despite encouraging wider AI adoption across the company. According to a Bloomberg report, the iPhone maker has lowered the budgets for employees to access AI tools and services to keep costs under control. It reportedly reduced the daily AI usage budget for some engineers from $300 to $200 last month.

Bloomberg further highlighted that Apple is considering additional reductions for some engineering teams, potentially bringing their daily AI spending allowance down to $100. Non-engineering teams could face even tighter limits, and some employees already have a daily AI spending cap of $100. This could potentially be reduced to as little as $25.

Must read: Deepinder Goyal says Temple beats Apple Watch Series 12, but urges some buyers to wait

Employees who need a higher AI usage limit could request a higher spending limit. However, these exemptions would require approval from senior executives. But how is Apple calculating the AI budget? Well, AI budgets are calculated based on AI tokens. Tokens are small units of text required for an AI model to process and respond. The use of AI tools such as ChatGPT, Claude, and Gemini incurs token-based costs, which can rise with increased usage and more computationally demanding requests.

However, Apple is not the only tech company reducing AI budgets for its employees. It has become an industry-wide trend, with companies like Accenture, Microsoft, Meta, Goldman ⁠Sachs, and others having significantly reduced AI adoption as part of cost-cutting.

Must read: Apple Intelligence is eating 15GB on your iPhone. Here’s how to cut it to 100MB with iOS 27 update

The trend highlights the growing concern between AI enthusiasm and real-world ROI. While tech companies are promoting AI as a transformative force, they are also increasingly scrutinizing compute costs, forcing engineering managers to justify every API token consumed. However, it does not mark the end of corporate AI deployment, but suggests a transition into more disciplined AI usage.