Apple is cutting daily AI spending limits for engineers

Summary

According to Bloomberg, Apple is reducing the daily AI spending limit for engineers from $300 to $200. Reports indicate the company plans to lower this limit to $100, or even as low as $25, for non-engineering units. This move follows the reality of rising token bills clashing with the industry-wide, short-lived 'tokenmaxxing' (maximizing token consumption) trend.

Why it matters

Apple is one of the best-resourced engineering organizations on the planet. The company's internal AI usage serves as an indicator of how major tech firms are integrating LLMs (large language models) into their daily workflows. Retreating from generous AI budgets suggests that the promised efficiency gains (ROI) have yet to be proven at scale. This development will have wide-ranging implications, from vendor pricing and corporate AI procurement to the 'AI-first' narrative shaping tech strategy in 2025-2026.

Highlights

• Engineering limits were reduced from $300 to $200 in October; reports suggest further cuts down to $100 are ongoing
• Daily AI limits for non-engineering staff could drop as low as $25
• This change follows the short-lived 'tokenmaxxing' push where companies encouraged maximum AI usage before bills arrived

Critical take

The article is a weak re-report based on a single Bloomberg source without any confirmation from Apple. Framing the situation as a 'short-lived frenzy' risks presenting a routine budget reallocation as a dramatic reversal. The lack of internal data on actual usage, output quality, or cost-per-task means that while Apple is reducing token spend, it says nothing about whether work has become cheaper, faster, or worse.