Apple has begun scaling back the daily budgets allocated to its employees for accessing artificial intelligence tools, a move driven by the escalating costs associated with using services like ChatGPT, Claude, and Gemini.
According to reports, the iPhone maker has lowered the daily AI spending limit for some engineers from $300 to $200. Further reductions are under consideration, potentially bringing engineering team allowances down to $100 per day. Non-engineering teams could face even tighter restrictions, with some employees already capped at $100 and potential reductions to as little as $25.
The Rising Cost of AI Tokens
The core of the cost issue lies in how AI budgets are calculated: based on 'tokens.' Tokens are small units of text that AI models process to understand queries and generate responses. Increased usage and more complex, computationally demanding requests directly translate into higher token consumption and, consequently, higher costs for the company.
While Apple is encouraging wider AI adoption internally, the financial implications of widespread, unrestricted access to these powerful tools have prompted a reevaluation of spending. Employees requiring higher AI usage limits can request exemptions, but these now require approval from senior executives.
An Industry-Wide Trend Towards Disciplined AI Spending
Apple is not alone in this cost-cutting measure. This trend is emerging across the tech industry, with major players like Accenture, Microsoft, Meta, and Goldman Sachs also reportedly implementing significant reductions in employee AI adoption budgets.
This shift highlights a growing tension between the enthusiastic embrace of AI's transformative potential and the practical need for a measurable return on investment (ROI). Companies are scrutinizing compute costs more closely, compelling engineering managers to justify every API token consumed. Rather than signaling an end to corporate AI deployment, this trend suggests a transition towards more disciplined, cost-conscious, and strategically managed AI usage within large organizations.