Overview
- In October Apple lowered daily AI spending for some engineers from $300 to $200, a change confirmed by multiple reports.
- Company discussions now include deeper cuts that could drop some engineering teams to $100 per day and certain non‑engineering teams to $25, with employees able to request exemptions from senior leaders.
- The allowances pay for AI “tokens,” the units models use when handling requests, and Apple lets staff access services such as OpenAI’s ChatGPT, Anthropic’s Claude and Google’s Gemini.
- Managers are directing teams to route simple tasks to cheaper models or run inference on devices to curb token bills, but savings are uncertain because many staff did not use full previous limits.
- The move follows similar cost controls at other firms and comes as Apple’s R&D and infrastructure spending rose, raising questions about whether token consumption has delivered clear productivity gains and prompting workflow and approval changes for employees.