Apple set to cap employee AI use amid rising costs

Apple is cutting back AI budgets for some of its teams, according to multiple reports.

The reports come as multiple tech companies and major consultancies have instituted their own AI token caps for staff.

Bloomberg reported via unnamed sources that the tech giant was considering rolling back employee budgets for AI as part of wider efforts to cut costs

Some engineers have already had budgets reduced from $300 to $200 last month, per Bloomberg, with further plans expected to cap AI token allowances to $100 a day.

TechCrunch echoed that report, adding that Apple was considering reducing the allowance for AI tokens to $25 a day for employees outside of engineering, from $100 now.

ITPro approached Apple for comment, but did not receive a response by the time of publication.

Apple opts for cheaper models

If true, the spending cut would impact tokens used to access third-party AI models, such as Anthropic's Claude or OpenAI's ChatGPT. The unconfirmed reports suggest that some teams were being told to use cheaper models or locally hosted versions to cut costs.

According to Bloomberg, employees who need a beefier allowance to complete their work will be able to request more tokens from management.

However, sources told the publication that many Apple staff already aren't making use of the existing limits, so cutting back on allocations could be more about releasing funds budgeted for employee token use rather than reducing AI use or cost cutting, noted AppleInsider.

Bloomberg suggested that some staff see the cap as a sign of trouble at the company, raising concerns about the threat of potential layoffs.

Apple earlier this week released its first quarterly results under new CEO John Ternus, with $114.5bn in fourth quarter sales, beating analyst estimates.

However, shares fell after a conference call with analysts revealed Apple's continued concerns with memory shortages driving up prices and potentially cutting in margins.

The end of tokenmaxxing?

If indeed true, the move comes amid a wider crackdown on so-called tokenmaxxing, the practice of using AI as much as possible that followed a rise in prices earlier this year, as well as concerns that AI isn't delivering on its promise to boost productivity and bottom lines.

A report from Cloudera earlier this year suggested nine-in-ten organizations have scaled back or reduced AI use due to the rise in token costs.

In June, Uber introduced a cap after chewing through its annual AI budget in just four months, limiting staff at $1,500 in tokens a month. The same month, Accenture followed suit by telling staff to stop using AI for unnecessary tasks such as converting PDFs to slides amid "soaring token spend".

In August, Microsoft updated internal guidance, advising staff to avoid tokenmaxxing and instead focus on creating benefits for customers and the business — though Microsoft added that it was still "AI first" and there was no formal cap.

Just last week, Amazon issued a similar update, telling its own software engineers that not every task will need AI, noting: "if your work does not benefit from current AI tooling, that is a finding, not a failing."

FOLLOW US ON SOCIAL MEDIA

Follow ITPro on Google News and add us as a preferred source to keep tabs on all our latest news, analysis, views, and reviews.

You can also follow ITPro on LinkedIn, X, Facebook, and BlueSky.