The Two Teams Bloomberg Named on the Same Day
On August 21, 2026, Bloomberg's Mark Gurman reported that Apple was cutting more than 200 jobs, and within hours Apple confirmed the cuts on the record instead of letting the report stand unanswered. A company spokesperson told Bloomberg that Apple is "looking to evolve our business to deliver the best experiences for our users," and that the changes "will also impact a limited number of existing roles."
The two teams named were not obscure cost centers. One built Vision Pro, the headset Apple spent years positioning as the start of a new computing platform. The other built Siri and the Intelligent Systems Experience group behind it, the voice assistant Apple has been racing to rebuild into something genuinely AI-capable. Cutting either team alone would read as routine. Cutting both on the same day, and confirming it in writing, is Apple telling the market exactly where it is willing to bleed to fund what comes next.
A Roughly Even Split: About 100 and About 100
The reporting that followed Gurman's story, corroborated separately by TechCrunch, 9to5Mac, MacTech, Macworld and AppleInsider, converged on a roughly even split: about 100 roles cut from the Vision Pro hardware and software group, and about 100 from the Siri and Intelligent Systems Experience team, with some gaming-adjacent roles affected as well. Apple's own statement named neither team and offered no figures beyond what the reporting had already surfaced.
| Group affected | Reported cut | Source |
|---|---|---|
| Vision Pro hardware and software | about 100 roles | Bloomberg (Gurman), corroborated by TechCrunch, 9to5Mac, MacTech, Macworld, AppleInsider |
| Siri / Intelligent Systems Experience | about 100 roles | Same corroborating reports |
| Total confirmed | 200+ roles | Apple, on-record statement to Bloomberg, August 21, 2026 |
An even split across two teams that could not be more different in status tells its own story. Siri is the project nearly everyone agrees needs more resources, not fewer, after years of being the industry's punching bag on AI. Vision Pro is the project Apple spent years insisting was not a punching bag, it was the future. Trimming the second to feed the first is not the cut most people expected first.
Why Cutting a Visible Winner Is Harder Than Cutting a Failure
The hardest reallocation call an owner ever makes is not shutting down a project that is obviously failing, because almost nobody defends that project very hard and the numbers usually make the call for you. The hard call is trimming a project that still has real users, real press coverage and real prestige attached to it, in order to fund something that has not shipped yet. Vision Pro still exists, still ships, still gets reviewed. A rebuilt, AI-capable Siri and Apple's reported smart-glasses push are still promises. Apple chose to shrink the thing that already works to feed the things that do not exist yet.
That is exactly the decision most owners avoid, because it is asymmetric in a way that flatters inaction. Protecting Vision Pro's headcount costs nothing today and only shows up as a missed opportunity later, quietly, in a quarter when a rival's AI assistant or a competitor's smart glasses arrive first. Cutting Vision Pro's headcount costs something visibly, this week, in a Bloomberg story about a flagship product. Sunk-cost momentum always favors protecting the visible thing and starving the invisible one, because only one of those costs ever shows up on a calendar anyone can point to.
The Test: Would You Fund This Project Today, at This Size
The test Apple's move suggests is not whether a project succeeded, it is whether you would start it today, at its current size, with what you now know. Vision Pro shipped, sold to real customers, and generated years of coverage; by the standard of "did it fail," it is defensible. By the standard of "would I staff this at this scale if I were deciding fresh today," Apple's own actions answer no. That is the threshold worth applying before your next budget cycle, not after a rival forces the question: pick the one team in your operation whose success is not in doubt, and ask whether its current headcount is a decision you would make fresh today, or one you are only protecting because you already made it once.
Run that test on the project that still gets good reviews, still has a dedicated internal champion, and still gets funded every cycle without a fresh argument. If the honest answer is that you would staff it smaller starting today, the gap between that number and its current headcount is not stability. It is sunk cost wearing the costume of a strategic bet.
Read next: The Yield Number Didn't Kill Apple's Anniversary iPhone After All | ChatGPT Can Now Send Texts From Your Work Mac
