Apple Cuts AI Tool Spending Limits for Engineers From $300 to $200 a Day
Apple’s engineers just had their AI budget trimmed. According to a new Bloomberg report carried by Gadgets 360, the company has cut the daily spending limit on AI tools for some of its engineers from $300 to $200 — and the ceiling could fall further still, to $100 a day.
On paper, $100 a day does not sound like much to argue about. But stretch that across thousands of engineers using AI coding assistants every day, and the numbers get serious fast. AI tools went from a novelty to an everyday cost in a matter of months, and Apple’s finance team has clearly started counting.
Here is what we know, why it is happening, and what it says about where Apple’s AI efforts stand.
What changed — and who it affects
The report says Apple has lowered the daily spending cap on third-party AI tools from $300 to $200 for certain engineers, with an internal discussion about taking it down to $100. Apple declined to comment on the report.
That possible $100 figure matters: it tells you this is not a one-off tweak. Apple is weighing how far the cuts should go, which suggests the spending has become a live question inside the company, not something decided once and filed away.
Engineers have also been nudged toward cheaper models that run locally instead of the priciest frontier models for every task. The idea is common sense — use the premium model where it earns its keep, and the cheaper option everywhere else — but it is also a sign of a company tightening its belt in a technology it is still trying to catch up in.
Why now?
Apple’s spending on research and development surged 32 percent in fiscal 2026, which ended in September, as the company works to catch up in the AI race. Hiring, compute, and access to outside models all add up, and the bill is now big enough to show up in budgeting decisions.
Some of that money is going toward an AI search tool the company is reportedly developing, and custom chips to power its own AI servers. Apple has long designed its own silicon for iPhones and Macs; now it is trying to do the same for the data centres behind its AI features — partly to keep costs down, and partly to stop depending on outside suppliers.
The timing is awkward, though. Apple is still seen as trailing rivals whose AI products ship faster, and it has only recently moved to deepen its AI partnerships in China. Cutting tool budgets while playing catch-up is a hard sell internally, which is probably why the cap has moved in steps rather than all at once.
What it says about Apple’s AI race
Here is the tension. Apple’s own revenue is booming — the company posted a record $114.5 billion in quarterly revenue — so this is not a company in trouble. It is a company choosing where its money goes. And right now, one of those choices is: less on AI tools for engineers, more on building its own AI infrastructure.
That is a defensible strategy. Running cheaper local models costs less than renting the most expensive model for every task, and in-house chips should pay for themselves over time. But the move also shows the limits of the moment: even Apple, with its enormous cash reserves, is watching the meter on AI.
And what about the engineers themselves? Their daily tools matter for output, and a $100 cap gives far less room to run heavy models than $300 did. Apple’s bet seems to be that better discipline — the right model for the right job — can recover the difference. We will see.
The bigger picture
Apple is not alone in asking hard questions about AI costs. The whole industry is discovering that AI is not just a research project; it is an electricity bill, a compute bill, and a software bill, all rolled into one. One of the reasons the market reacted sharply to Big Tech’s AI spending this year is that the sums involved are enormous — and the returns are still being measured.
Against that backdrop, Apple’s cap looks less like a quirky internal memo and more like a signal: the age of unlimited AI spending is over, and from here on, every company has to show what the money buys.
What to watch next
Watch whether the cap lands at $100 or stops at $200, and whether the cuts spread beyond the engineers named in the report. Watch too for Apple’s AI search tool and the custom server chips, both part of the same drive to bring costs down.
And watch the wider industry. If Apple’s engineers learn to build with cheaper models and in-house hardware, other companies will follow. The report behind this story is short on drama, but it points at the question every AI company will have to answer: how much are you willing to spend for smarter software, and where do you draw the line? Apple’s own site carries the company’s public announcements and product pages.