Apple Cuts Employee AI Usage Budgets, Some Teams' Daily Limits Reduced to $100
nashnova research
Apple has quietly lowered some engineers' daily AI-tool spending cap from $300 to $200 and plans to push it to $100, while non-engineering teams may see their limit slashed to $25 — even as the company's overall AI spending surges by double digits.
How deep are the cuts?
Engineers' daily cap: $300 → $200, with some teams headed to $100.
Non-engineering teams already face a $100 daily limit; Apple is considering cutting that to $25.
Staff who need more must request a waiver from senior management. This means → routine AI use shifts from "tap and go" to "file and wait."
What are tokens, and why does the cap sting?
A token — the smallest unit of text or data an AI model processes per request — is how AI providers bill usage. Stronger models and complex tasks like coding burn through tokens faster.
Apple employees currently have access to Anthropic's Claude, OpenAI's ChatGPT, and Google's Gemini.
Some teams have already been told to switch to cheaper models or run AI locally on-device. In plain terms = moving from first class to economy — or driving yourself.
How are employees reacting — and will the savings matter?
Reactions are split: some say few people ever hit the $300 cap, so the old limit was already generous; others read the move as a broader cost-cutting signal and worry about layoffs.
Actual savings remain unclear — many employees never came close to the old ceiling. This means → the current cut lands more on psychology than on real spending.
If deeper reductions go through, heavy users will have to request waivers repeatedly, and the drag on productivity will become harder to ignore.
Employee budgets are shrinking — so why is Apple's total AI bill still growing?
Apple disclosed this week that FY2026 R&D spending rose 32% year-on-year, driven mainly by infrastructure and AI costs.
Company-wide operating expenses jumped 22% to $75.5 billion, or 16% of total net sales.
This reflects a striking contradiction: the company is pouring money into AI at the corporate level while asking individual employees to spend less. In plain terms = the money hasn't shrunk — it's just going somewhere else.
Is Apple the only company doing this?
No. Microsoft has tightened internal AI spending limits. Tesla caps employee AI use at $200 per week. Uber limits a single AI coding tool to $1,500 per month. Meta has introduced token budgets and is steering staff toward in-house AI tools.
This reflects a sector-wide dilemma: encouraging deep AI adoption while controlling runaway usage costs — and no one has found the balance yet.
New CEO John Ternus is pushing Apple to run faster and leaner; some divisions have brought in outside consultants to assess efficiency. Whether cutting AI budgets can reduce costs without eroding the productivity gains AI delivers is the ultimate test of this balancing act.
市场有风险，内容仅供研究参考，不构成投资建议。