Apple cut Siri engineers while Siri trails ChatGPT. That is not a rebuild strategy

Apple‘s statement describing the restructuring used the language of momentum: the company is working to “evolve our business to deliver the best experiences for our users.” The cuts affect two groups — roughly 100 employees in the Vision Products Group and another 100 in the Intelligent Systems Experience team — and Bloomberg, TechCrunch, and MacRumors all confirmed the scope within hours of each other. The announcement came with the requisite reassurance that new roles would be created. What the statement did not explain is why reducing the team responsible for Siri’s AI features is the right move at the moment when Apple’s AI assistant trails its main competitors by the largest margin in the product’s history.

The Vision Pro side of this story is, on its own terms, defensible. Three years after Apple declared the dawn of spatial computing, the $3,499 headset has not broken through to any meaningful consumer base. By late 2025, quarterly unit sales had dropped to approximately 45,000 — a figure that made the device among the least-purchased Apple products in recent history. At that adoption rate, the economics of producing original immersive video content — which, according to Bloomberg’s reporting, costs millions of dollars per episode to film — were impossible to justify. Apple shuttered the gaming team and scaled back the immersive video unit. This is the kind of correction a rational company makes when the numbers stop lying to it.

The logic behind those cuts extends further than just the headset’s sales figures. App Store data for visionOS has consistently shown developers gravitating toward utility applications — virtual desktops, productivity environments, enterprise tools — rather than gaming. The immersive gaming potential Apple promoted at launch remained largely theoretical even two years in. Third-party studios were always going to produce this content more efficiently than an in-house unit whose production costs were never reconcilable with a device that most consumers have never held. Shutting down those teams and shifting to a third-party developer ecosystem is the correction those numbers demanded.

Apple is also making a platform bet that is becoming clearer. Smart glasses — lighter, cheaper, designed for daily wear — have moved up the internal priority list, according to TechCrunch and Digital Trends, with a target release of 2027. A second-generation Vision Pro is now positioned for 2028, after an overhaul aimed at enterprise. This trajectory makes more sense than the original consumer-spatial-computing pitch. The Vision Pro in its current form is a device for a particular kind of enterprise user; smart glasses are the device that might actually reach the market the headset could never find.

Apple’s defenders have a coherent case for the Vision Pro cuts specifically. The company has done this before — killed the AirPower mat, abandoned the Mac Pro trash-can design, retired the iPod Touch, moved on from every failed form factor without much ceremony. Apple routinely stops doing things that do not work. In this reading, the gaming and content cuts fit a pattern: Apple tested a hypothesis, collected the data, and is now reallocating. The analysis is not embarrassing — it is characteristically rational.

The same argument applies, tentatively, to the Siri cuts. The people being let go from the Intelligent Systems Experience team are, in Apple’s framing, engineers working on legacy approaches to AI features — not the people building the new Siri that Apple has been announcing since 2024. A standard tech-company restructuring around a platform transition reduces the old team while creating new roles for the new direction. Apple created Siri in 2011 and has been rebuilding it, in stages, for years. Cutting 100 people from a team doesn’t necessarily mean cutting capacity — it may mean moving that capacity toward an architecture that can actually compete.

The problem is the gap between the hypothesis and the evidence.

Apple’s rebuilt Siri, in the specific form promised at WWDC 2024 and reiterated the following year, is not yet the product Apple has described. The iOS 27 public beta shipped with AI features that arrived with an asterisk — several capabilities flagged in Apple’s announcements were still absent or limited as of the beta’s release. This is not an isolated delay in a complex product cycle; it is the second consecutive year that the rebuilt Siri has been the most heavily marketed feature Apple is not yet delivering.

The competitive gap is now concrete and public. ChatGPT answers roughly 30% more questions than Siri and does so with approximately 25% greater accuracy, according to comparative benchmark analysis published by technology observers. Google‘s Gemini has been integrated into Apple’s own operating system — an arrangement that is easy to read: when your own AI assistant cannot meet user expectations, you pay your most capable competitor to fill the gap on your behalf. Apple is collecting revenue from Google for AI traffic on iPhones that Apple’s own assistant cannot handle. That is not a minor or temporary condition. It is the operating reality of the world’s most valuable company’s most important software product.

The lawsuit Apple filed against OpenAI in April 2026, alleging trade secret theft, adds another layer to the picture. Apple is simultaneously suing the company whose technology it is implicitly competing against, partnering with a competitor’s AI to fill Siri’s gaps, and cutting 100 engineers from the team responsible for AI features on devices used by more than 2 billion people. These are not contradictory — but only if the internal rebuild is further along than any external evidence suggests. Apple has earned the benefit of the doubt on platform transitions more than almost any company in the history of consumer electronics. That credit is not infinite.

The Vision Pro and Siri situations differ in a specific way that matters for how to read these cuts. The Vision Pro gaming team was building something a third party could also build — and, given the right device adoption curve, would probably build better. Content produced internally at the cost of millions per episode is bad economics for a device with 45,000 quarterly sales; the same content produced by an independent studio, for a larger and more engaged platform, is a different calculation entirely. The in-house team was not essential to Vision Pro’s future. Cutting it did not reduce the headset’s potential.

Siri’s situation is the inverse. Apple’s AI assistant is not a feature that can be handed to a third party and remain Apple’s — it is the connective tissue of how the company’s devices understand what users want, anticipate their next move, and handle requests across the operating system. You cannot delegate Siri’s development without Siri ceasing to be Apple’s product. The partnership with Google is a workaround, not a feature. Reducing the Intelligent Systems Experience team while that workaround is still load-bearing is, at minimum, a decision that requires more explanation than “we are creating new roles.”

Apple’s development process — design privately, announce at WWDC, ship in September, iterate deliberately — was built for hardware-driven innovation cycles. It is not well-suited to the pace at which language model capabilities are advancing. Competitors ship meaningful AI capability updates on cycles measured in weeks. Apple ships annually. When the product is a physical device — a chip, a screen, a camera system — the annual cadence works because manufacturing has inertia of its own. When the product is a software capability that a competitor can update with a model push, the annual cadence is a structural disadvantage.

What is established — and where the argument starts

What is confirmed: Apple cut more than 200 jobs across Vision Pro gaming, immersive video, and Siri and AI teams in August 2026, as reported by Bloomberg, TechCrunch, and multiple other outlets. Apple’s statement confirmed the cuts and framed them as strategic evolution. Vision Pro’s immersive content production had become economically unsustainable at current device adoption rates. Siri trails its main competitors on accuracy and question coverage by measurable margins. None of this is disputed.

What is not settled: whether the Siri cuts remove legacy engineering capacity that no longer fits the new architecture, or whether they reduce the capacity Apple needs to close the gap it has been promising to close for two years. Apple’s assurance that new roles will accompany the cuts is consistent with a genuine platform rebuild. It is also consistent with the language every major tech company uses when reducing headcount. The distinction matters, but it is not visible from outside Apple’s internal roadmap. What is observable is the gap between the AI capabilities Apple has announced and the AI capabilities users can access today — and that gap is not narrowing faster than the competition is advancing.

The Vision Pro gaming and content cuts are, by any reasonable reading, the right call two years late. Apple is acknowledging that it cannot produce the consumer entertainment ecosystem its headset requires from the inside, and it is adjusting to focus the device on the enterprise users who have actually shown up. That is clear strategy, even if it took 45,000 quarterly unit sales to become undeniable.

The Siri situation is not clear. A company that is two years into a promised rebuild of its AI assistant, visibly unable to close the gap with competitors that ship updates on a schedule Apple does not match, has cut 100 engineers from the team responsible for that assistant. The new roles that will be created may be the right roles. The direction may be the one that eventually closes the gap. But the gap is here now, the competition is moving now, and cutting the team is the part of this announcement that the rebuilt Siri, whenever it arrives, will have to answer for.

Tags: artificial intelligence, Apple, Siri, Apple Vision Pro, layoffs