Apple has reduced its internal AI spending by lowering the daily token budget for engineers from $300 to $200, with further reductions to $100 or as low as $25 planned for certain non-technical teams. Staff access to external services like Anthropic’s Claude, OpenAI’s ChatGPT, and Google’s Gemini remains available, though some groups will switch to cheaper models or tools running locally on devices. This decision aligns with CEO John Ternus’s push to make the company leaner, even as research and development spending rose 32 percent in fiscal year 2026 due to higher infrastructure costs. Reactions among employees are mixed, with some viewing previous limits as excessive while others fear these measures could signal upcoming layoffs. Tech firms including Microsoft, Tesla, Uber, and Meta have also recently capped internal AI expenditures.
The shift reflects a broader industry trend where companies prioritise cost control over unrestricted access to advanced models during internal work. It suggests that organisations are moving away from relying heavily on external APIs for routine tasks, potentially favouring on-device processing or smaller, more efficient models. This approach may reduce long-term operational expenses without necessarily altering the capabilities of customer-facing products or services.
- Daily token budgets now range from $200 down to $25 depending on team role.
- Staff must request exceptions from management for higher usage limits.
- Several major tech firms have adopted similar spending caps recently.
