Apple cuts daily AI spending limit and re-evaluates generative tools usage

1–2 minutes

AI budget cuts for engineers

Apple has reduced the daily spending limit on AI models for some of its engineers from $300 to $200, a one-third cut implemented in October, according to Bloomberg. Further restrictions are under discussion, which could result in a $100 daily quota for certain engineering groups. Non-technical teams already operate under a $100 cap, which could be reduced to $25 in specific cases.

Pressure on R&D costs

In fiscal year 2026, Apple’s R&D investments reached $45.8 billion, a 32% increase from the previous year. The company attributed this growth primarily to infrastructure spending, including AI initiatives, and personnel costs. Total operating expenses rose to $75.5 billion, a 22% increase, accounting for 16% of net sales.

Impact on employees’ daily work

Apple employees have access to services like Anthropic’s Claude, OpenAI’s ChatGPT, and Google’s Gemini. Following the budget review, some teams are being instructed to prioritize cheaper models or use locally run versions to control costs without halting projects.

Employees requiring additional resources can request exceptions from senior leadership. Apple declined to comment on Bloomberg’s report.

Market context and lessons from other companies

Apple’s move reflects a challenge faced by other large companies. For instance, Uber imposed limits on AI tool usage after exhausting its annual budget for the technology within months. The widespread adoption of generative models has forced managers to determine where investing in more expensive solutions is justified, given the rising costs.

Balancing innovation and cost control

As Apple increasingly integrates AI into its ecosystem, particularly through Siri and other devices, strict cost control is essential to maintain financial health. The reduction in daily spending limits and the push for more cost-effective alternatives demonstrate the company’s efforts to balance technological ambition with budgetary constraints.