Your Mac Has a New Update, and Siri Can Do More
Apple released macOS 27 Golden Gate on September 14. The Verge It arrives on the same date as iOS 27, iPadOS 27 and watchOS 27. Engadget
It runs only on Macs with Apple Silicon. These are newer Macs with Apple's own chips. Older Macs with Intel chips cannot install it.
The main addition is the new Siri AI assistant. It ships as a beta, an early test version. At launch it works in English only. A separate Siri app saves your past chats.
Siri AI can search your own material, such as photos and emails. It can also do tasks in apps, including Messages and Reminders.
There are also small changes to how things look. Windows now use the same rounded corners. Liquid Glass, the clear layered look for buttons and panels, has steadier lighting and clearer text. A slider lets you make those parts more solid or more see-through.
Staged rollout, caps and regional split
Apple Intelligence and Siri AI across iOS 27, iPadOS 27, macOS 27, watchOS 27 and visionOS 27 need iPhone 16 models or later. Apple On Mac, Siri AI is a beta for people with a supported device set to English.
Siri AI will still be in beta when the OS 27 versions reach the public on September 14. AppleInsider At launch there are daily limits on use.
Rules differ by device in Europe. Apple will not ship Siri AI in the European Union with iOS 27 and iPadOS 27 because of the Digital Markets Act. People in the EU will get Siri AI on macOS 27 and visionOS 27. Apple
Developers had early access. New Siri AI features have been open for developer testing in iOS 27, iPadOS 27, macOS 27 and visionOS 27. Testing for watchOS 27 comes later, in a future watchOS 27 beta.
watchOS 27 needs a recent watch. It will run on Apple Watch Series 9 or later, Apple Watch SE 3, and Apple Watch Ultra 2 or later. Apple
What matters for builders and fleets
The broader context here is Apple keeping demand low while it learns. English-only support, limits to newer devices, beta status and daily limits reduce use while Apple studies real use. That care makes sense for a helper that reads photos and email and acts in other apps. Early reports on speed and reliability will mostly come from English-speaking users on newer hardware.
Looking at what this means for app makers, testing gets harder. Apps that use Siri AI will need a backup plan for three cases: older devices that lack support, EU iPhones and iPads with no Siri AI at launch, and users who hit a daily limit mid-task. Because Siri can read personal material and act in Messages and Reminders, clear permission requests, tight limits on prompts, and clear records will matter more. The chat history should be treated as saved data users will want to search, export and delete.
Looking at what this means for workplaces with many Macs, the Intel cutoff cuts both ways. Apple Silicon only means one type of chip to test for OS 27, one set of expectations for on-device work, and a clearer plan for retiring Intel machines still in use. Machines that cannot upgrade will not get the new assistant, the Liquid Glass updates or the window fixes, which will bring upgrade plans forward.
In my view, the Siri app is the change to watch. Siri moves from a short voice reply to a saved, searchable record of work. I have watched my own adult children use chat histories as filing cabinets, searching them long after. If that habit moves to Siri, finding the right photo or email, and showing clearly what was done in Messages and Reminders, will matter more than smooth replies.
Another point worth flagging is the split in Europe. Having Siri AI on Mac and visionOS in the EU, but not on iPhone and iPad, creates a gap for people who switch between Mac and iPhone during the day. For developers with Apple Intelligence access, checking by system and region will be required. The coding is workable. The harder job is explaining to users why the helper is on one screen and missing on another.
Looking further ahead, these limits point to what comes next. Limits control computing cost. Language limits control testing cost. Hardware limits reduce differences in on-device speed. Once those settle, wider language support and higher limits become practical choices. Past moves from on-device to cloud-assisted computing suggest early limits can make room for steady everyday use later.
