Apple reportedly cuts some engineers' daily AI budget from $300 to $200

Earlier this year, many tech companies urged their employees to use artificial intelligence as much as possible. At Apple, rationing is reportedly now the order of the day. In October, the company is said to have lowered the daily AI spending cap for some of its engineers from $300 to $200. The information comes from Bloomberg and was picked up on November 4 by TechCrunch. Apple did not respond to TechCrunch's requests for comment.

The cuts may not stop there. According to Bloomberg's sources, some budgets could fall as low as $100 per day. Outside engineering, where many teams are already capped at $100 per day, Apple reportedly considered going as low as $25 for some of them. TechCrunch's report does not specify which tools are affected or whether these amounts apply per person or per team.

After the race to consume, the bill

TechCrunch places the decision in a broader trend. Earlier this year, many companies pushed their employees to use AI heavily, in the name of poorly defined efficiency gains. This race to consume “tokens,” the units of text that AI providers bill for, was nicknamed “tokenmaxxing.” “The frenzy was short-lived,” the site writes: when the bills arrived, companies started rationing, an approach seen as more financially sensible.

In the background, TechCrunch points to the gap often noted between what AI costs, both to build and to use, and still unclear productivity gains. Torsten Slok, chief economist at the asset manager Apollo, recently estimated that outside the tech sector, AI is not yet paying off for American companies.

How to avoid having to cut abruptly

If Apple, one of the richest companies in the world, is counting its AI dollars, an SME has every reason to do so from the start. A few good habits:

• Set quotas per user or per team, with an alert before the cap is reached, rather than an unlimited budget that will have to be cut later.
• Track consumption by use case (programming, writing, document analysis) to know where the money goes.
• Reserve the most expensive models for tasks that justify them and use lighter models for the rest.
• Measure what the tool delivers: time saved, requests handled, code shipped. Without a metric, the budget will be impossible to defend the day the bill climbs.

It remains to be seen whether Apple will confirm these cuts. According to Stocktwits, cited by Yahoo Finance, Microsoft, Tesla, Uber and Meta have already tightened control over their internal AI tools. If rationing becomes widespread, AI providers, which count on rising corporate consumption, could feel the impact.