Apple Slashes Employee AI Budgets as John Ternus Cuts Costs: Report

People familiar with Apple’s internal AI spending say the company has cut daily allowances for some engineers and is planning further reductions even as CEO John Ternus pushes Cupertino to run faster and leaner.

Citing people familiar with the matter, Bloomberg reports that Apple last month lowered the daily spending cap for some engineers using AI models to US$200 from US$300.

Those sources tell Mark Gurman the company is now planning further cuts for some engineering teams, potentially reducing their daily allowances to as little as US$100. The people asked not to be identified because the discussions are private.

Many groups outside of engineering already face a US$100 daily cap, but Apple has considered reducing that limit to as little as US$25 for some teams, the people said. Employees who need higher allowances for their jobs will be able to seek exemptions from senior leadership.

The budgets cover tokens, the units of text and other data an AI model processes when it handles a request. Providers typically charge based on how many tokens are used, with more advanced models generally costing more, and complex coding work can burn through large amounts.

Reactions inside the company have been mixed, according to the people familiar with the changes. Some employees say few workers were using their full US$300 daily allowances, so the previous limits looked unnecessarily generous. Others see the reductions as a sign of broader cost cutting and worry about possible layoffs.

Already, some Apple divisions have hired outside consultants to evaluate efficiency and operations, fuelling employee anxiety about possible restructurings. Ternus became CEO on September 1 and has been pressing for fewer management layers alongside the leaner approach.

Savings from the AI cap cuts remain unclear, especially because some employees were not approaching the earlier limits. Heavier users could be forced to seek exemptions if further reductions go ahead.

Overall spending at Apple has still risen. The company said this week that research and development expenses increased 32 per cent in fiscal year 2026, driven by higher infrastructure and AI costs, while companywide operating expenses jumped 22 per cent to US$75.5 billion, or 16 per cent of total net sales, in the same fiscal year results Ternus delivered in his first report as CEO.

Staff already have access to Anthropic’s Claude, OpenAI’s ChatGPT and Google’s Gemini. Some teams are being directed to use less expensive models or to rely on AI that runs locally on devices, even as Apple continues to push AI into products such as Siri.

Other big tech firms are also tightening internal AI spending. Microsoft has reduced allowances, Tesla has imposed a US$200 weekly limit, Uber has capped individual AI coding tools at US$1,500 per month, and Meta has introduced token budgets while steering staff toward internal tools.