Every spring, a few weeks before WWDC, Apple previews the accessibility features it is shipping later in the year. It is one of the rare Apple announcements where the marketing and the real-world impact actually match up. This year the theme is clear. Apple's accessibility features are being rebuilt around Apple Intelligence, and several of them move from neat demo to something people will use every day.
The full list is long, so here is what genuinely changed, what it means in practice, and the parts the press release quietly buries (like UK availability).
What Apple announced
Apple previewed updates to VoiceOver, Magnifier, Voice Control and Accessibility Reader, all running on Apple Intelligence, plus on-device generated subtitles, an eye-controlled wheelchair feature for Vision Pro, and a wider rollout of the Hikawa adaptive grip. Most of the software lands later this year, which lines up with the usual pattern of a preview now and a proper reveal at WWDC on 8 June.
Tim Cook framed it around privacy, saying Apple is adding new capabilities "while maintaining our foundational commitment to privacy by design". That is the thread running through all of this: the AI work happens on device wherever possible, not in the cloud.
VoiceOver and Magnifier can describe the world around you
The biggest change is in how VoiceOver describes things. The new Image Explorer uses Apple Intelligence to give far more detailed descriptions of images across the system, not just photos but scanned bills, letters and other documents. Instead of a flat label, you get something closer to a paragraph of context.
Live Recognition also gets the Action button treatment. Point the camera, press the Action button, and you can ask what is in front of you and then ask follow-up questions in your own words. Magnifier gets the same ability with a high-contrast interface built for low vision, and you can drive the app by voice with commands like "zoom in" or "turn on flashlight".
If you have never explored what the Action button can already do, this is a good moment to. I have written about that and other hidden iPhone features Apple doesn't advertise, and accessibility settings are where a lot of the best ones live.
Voice Control finally speaks plain English
Voice Control is the update I think more people will notice, accessibility need or not. Until now it has leaned on numbered overlays and exact labels. With Apple Intelligence you can describe what you want in natural language, which Apple is calling "say what you see".
That means saying "tap the purple folder" or "tap the guide about best restaurants" instead of memorising a label. It works in visually dense apps like Maps and Files, and it can route around buttons that developers never labelled properly for accessibility in the first place. One caveat the release buries in a footnote: the natural-language version is English only at launch, in the US, Canada, the UK and Australia. The UK makes the list this time, which is not always a given.
Accessibility Reader handles documents that used to break it
Accessibility Reader is the customised reading view for people with dyslexia, low vision and related needs. The weakness has always been messy source material. The new version handles multi-column layouts, embedded images and tables, so scientific papers and PDFs no longer fall apart when you reformat them.
It adds on-demand summaries so you can get the gist before reading in full, and built-in translation that keeps your font, colour and formatting choices intact. That last detail matters more than it sounds, because losing your custom formatting on translate has always defeated the point.
Generated subtitles for any video, processed on device
This is the feature with the broadest reach. Apple is adding automatically generated subtitles for any video that does not already have them, including clips you record, things friends send you, and streamed content. It runs on on-device speech recognition, so the audio never leaves your device, and it works across iPhone, iPad, Mac, Apple TV and Vision Pro.
Sarah Herrlinger, Apple's head of accessibility policy, tied this to the same privacy theme, stressing that the processing is built to protect users at every step. The honest caveat: generated subtitles launch in English in the US and Canada only, so this is one UK users will be waiting on.
Driving a wheelchair with your eyes on Vision Pro
The most striking announcement is a power wheelchair control feature for Vision Pro. For people who cannot use a joystick, the headset's eye-tracking becomes the input method for compatible alternative drive systems. Apple's pitch is that the eye tracking does not need constant recalibration and copes with varied lighting.
It launches with the Tolt and LUCI drive systems in the US, over Bluetooth or wired, and is meant for controlled environments rather than open-road independence on day one. Pat Dolan of Team Gleason's advisory board, who has lived with ALS for a decade, said the ability to control his wheelchair himself is "gold to me". It is a narrow launch, but it is the kind of feature that justifies Vision Pro's existence far better than the entertainment pitch. I have been fairly critical of where Vision Pro is heading, and this is the sort of work that complicates that view in a good way.
The Hikawa Grip is the quiet star
The one thing you can actually buy today is the Hikawa Grip & Stand for iPhone, now available worldwide in three new colours. Designer Bailey Hikawa built it with disability communities involved from the start, for people whose grip, strength or mobility makes holding a phone difficult, and it is now a collaboration with PopSockets so it can ship globally.
It is a MagSafe accessory, which is worth noting given Apple is reportedly rethinking MagSafe on future iPhones. Adaptive accessories like this are a strong argument for keeping the magnets exactly where they are.
The smaller updates that still matter
A few more changes did not get headline billing but are worth listing:
- Vehicle Motion Cues come to visionOS to reduce motion sickness for Vision Pro passengers, plus face gestures and Dwell Control improvements
- Touch Accommodations get a new setup option in iOS and iPadOS
- Made for iPhone hearing aids pair and hand off between Apple devices more reliably
- Larger Text support arrives on tvOS for the first time
- Name Recognition, which alerts deaf and hard-of-hearing users when their name is said, now works in over 50 languages
- A new FaceTime API lets developers add a human sign language interpreter to a call
- The Sony Access controller can now be used as a standard game controller on iPhone, iPad and Mac
When you can actually use these
The pattern here is preview now, ship later in the year. Apple Intelligence remains in beta with a fixed language list, the natural-language Voice Control is English-only at launch (UK included), generated subtitles are US and Canada only for now, and the wheelchair feature is US only with specific hardware.
None of that is unusual for an Apple preview, and it fits the wider picture of Apple treating its AI rollout as a slow build rather than a single big swing. I have argued before that Apple's AI strategy is a marathon, not a sprint, and accessibility is the clearest evidence for why that approach can work.
Why this lands differently from the rest of Apple's AI news
Apple Intelligence has had a rough year of reception, much of it deserved. Accessibility is the exception. These are features where on-device AI has an obvious, concrete payoff: describing a bill to someone who cannot see it, subtitling a video with no captions, letting someone drive their own wheelchair. If you want the most convincing argument for Apple Intelligence so far, it is not Genmoji or writing tools. It is this.
FAQ
When will Apple's new accessibility features be available?
Apple says the Apple Intelligence updates and most of the new features arrive later in 2026, in line with its usual autumn software cycle. The Hikawa Grip & Stand is the exception and is on sale now.
Will the new Voice Control work in the UK?
Yes. The natural-language version of Voice Control launches in English in the US, Canada, the UK and Australia. Generated subtitles, by contrast, are US and Canada only at launch.
Does any of this need an internet connection?
The headline AI features are designed to run on device, including generated subtitles, which use on-device speech recognition so audio stays private. Some Apple Intelligence features still have hardware and language requirements.
Based on Apple's accessibility announcement on Apple Newsroom.
Lewis Lovelock
YouTuber, tech creator and CTO. I write about the apps, gear, and workflows I actually use — and make videos about them too.
Watch on YouTube →