Skip to content
D Consultants

Apple cuts internal AI budgets: how UAE businesses should set their own limits

Bloomberg reports that Apple has cut daily AI budgets for some staff. Apple has not confirmed the details. Here is what the shift means for UAE businesses, and how to control AI costs without slowing your teams.

Published 4 min readD Consultants team
An analog usage meter on a dark desk, with scattered dots of light flowing through narrow gates into neat, even rows
Share this article
Summarize with AI

Key points

  1. 01Bloomberg reports Apple cut some engineers from $300 to $200 a day in October.
  2. 02Further cuts are under consideration, and Apple has not confirmed the details.
  3. 03Some teams are being asked to use cheaper or on-device models.
  4. 04Other large companies have also set AI limits.
  5. 05UAE businesses should measure cost per task before setting caps.

Bloomberg reported on November 4, 2026 that Apple has cut internal AI budgets, with some engineers moved from $300 to $200 a day in October. If a company that size is putting limits on AI use, your business should decide its own limits before the invoice decides for you.

What changed

According to Reuters, citing Bloomberg, Apple is considering a further cut to $100 a day for some engineering teams. Some non-engineering groups already have a $100 daily cap, which could drop to $25. Employees who need more can ask senior executives for an exemption.

The budgets cover tokens, the units of data that models such as ChatGPT, Claude and Gemini process. Some teams are being asked to use cheaper or on-device models. MacRumors also reports employees telling Bloomberg that few people use the full $300, so the savings may be small.

Apple has not confirmed the details. TechCrunch sought comment and reported no response, and The Next Web says Apple declined to comment.

Why it matters for UAE businesses

Apple is not alone. The Next Web lists similar limits at Tesla, Uber and Accenture, and division-level budgets at Microsoft. Atlassian gives research and development staff monthly AI wallets set by role. Reuters also relays a Bloomberg report that Goldman Sachs restricted some AI coding tools after reviewing productivity and risk.

We draw three lessons from this.

First, AI is a metered cost. A flat subscription hides this. Usage-based tools and AI agents do not, because every task consumes tokens.

Second, a cap is a blunt tool. If few people reach the limit, the cap saves little. Apple's reported move to cheaper models may matter more than the number itself.

Third, the useful question is not how much you spend. It is what each task costs and what it returns. Most of the reports above describe limits on spending, not measured results.

If you run customer support, sales or back-office work through AI, this applies to you now. A cost problem is easier to fix while usage is small.

What to do now

Start with visibility, then set limits. Our AI and data team does this work, and we design AI agents and automation with cost per task in mind from the start.

Your next steps

  1. List every AI tool and bill

    Record each tool, owner, pricing model and monthly cost. Include personal accounts that staff expense.

  2. Measure cost per task

    Pick your three most-used workflows. Work out what one completed task costs in tokens or fees.

  3. Match the model to the job

    Send simple tasks to cheaper models. Keep the most capable models for work that needs them.

  4. Set budgets by role

    Developers and support teams use AI differently, so give them different limits. Add alerts before a limit is reached.

  5. Create a clear exception path

    Name who can approve extra budget and how fast. A limit should not stop urgent work.

Questions

Should a small business set daily AI caps?

A monthly budget by team is usually a better start. It is simpler to run and gives you room for busy days. Tighten it once you know your real usage.

Will spending limits slow our teams down?

They can, if the limit is set without data. Base limits on measured use, add alerts, and keep an exception path so urgent work continues.

Does this mean AI is not worth the cost?

No. The reports show companies controlling spend, not abandoning AI. Apple's own budgets remain, and the question is which tasks earn their cost.

Tags:AI costsAI strategyAI tokensAI agentsUAE business
Share this article

Written by

D Consultants team

Software and AI team, Dubai

The people who build and run dconsults.com, our own products atnos, Vline and blaaa, and websites and AI agents for businesses across the UAE.

Profile

More articles to explore

All articles

Tell us what you are building.

Free 30-minute call. We map your workflow, find the highest-impact opportunity and send a clear proposal.

+971 50 807 0826