Thursday, October 8, 2026
English edition

Development

macOS 27 Golden Gate: The Ars Technica review

September 16, 2026 Development Source: Ars Technica

macOS 27 Golden Gate: The Ars Technica review

Share this article

That means that the Golden Gate update requires an Apple M1 chip or better/newer, a list of hardware that includes the following: My primary testing hardware for this review has included an M1 MacBook Air (16GB/1TB), a 15-inch M3 MacBook Air (24GB/512GB), and a MacBook Neo (8GB/512GB), which means I have a pretty good sense of how Golden Gate and future macOS releases will treat Apple’s least-capable hardware. The good news is that if you own any Apple Silicon Mac that isn’t falling apart, Golden Gate should run mostly happily. It also means all Macs support Apple Intelligence, which never ran on Intel-based hardware. So far, the only features that require the more-capable model are more expressive voice capabilities for the new Siri and an improved voice-dictation feature. But I would expect to see the gap between older/less-capable Apple Silicon Macs and new ones grow, if and when Apple makes more changes to Apple Intelligence. Golden Gate is our second consecutive release named for a body of water, following a stretch of releases named for towns/regions (Big Sur, Monterey, Ventura, Sonoma, Sequoia) and another stretch named after things that are mostly rocks (Yosemite, El Capitan, Sierra, High Sierra, Mojave, Catalina). Maybe naming the Liquid Glass releases of macOS for bodies of water was intentional and maybe it wasn’t—only Craig Federighi can say for sure. You might think that macOS 27 is named for the Golden Gate Bridge since it’s an iconic landmark and it factors into all the real-world desktop wallpaper videos in the operating system (there’s also, per usual, an “abstract color swirl” option, which in this release is gold that fades to black). Since opening to the public in 1937, this bridge has become synonymous with San Francisco and the Bay Area, the same way that the Empire State Building symbolizes New York City, or that sports hooligans symbolize Philadelphia. But that’s not right! Did you know that the Golden Gate Bridge is so named because it’s a bridge that spans a strait called the Golden Gate? And not because it’s a bridge that is sort of arguably golden-looking and loosely resembles a gate? It had never once occurred to me to check my assumptions about this until now. The name dates back at least as far as the 1840s, according to the memoirs of explorer and politician John Charles Frémont, which say that he recorded the name on a map of the area that he submitted to the US Senate in 1848. The name was appropriate, given the gold rush that began swelling the area’s population in that same year. In an apples-to-apples comparison between Tahoe 26.6.2 with Apple Intelligence turned on and Golden Gate 27.0 beta 7 with Apple Intelligence turned on, Golden Gate needed a little over 4GB of extra space on an M1 MacBook Air, which is already a larger-than-usual increase. If you were running Tahoe with Apple Intelligence turned off, Golden Gate needed a whopping 17GB more space on the M1 MacBook Air, thanks to a mandatory multi-gigabyte language model download that kicks in automatically the first time you connect a freshly restored Golden Gate Mac to the Internet. And if you have a Mac that can run the AFM 3 Core Advanced model—remember, that’s any M3-or-newer Mac with 12GB or more of RAM—you’ll need even more disk space. The Storage tab of the System Settings app says that the M1 MacBook Air and MacBook Neo needed around 14GB of space for Apple Intelligence after a fresh install. On the M3 MacBook Air, that number was 22.42GB. One could be forgiven for being upset about a software update that consumes so much disk space in the middle of a historic RAM and storage shortage. What’s particularly frustrating is that Apple Intelligence is still fundamentally a fully optional component that is layered over the top of the core macOS experience. Downloaded installers and recovery files don’t include it; if you set up a new Mac and don’t connect it to the Internet, everything else about the Mac still looks and works much like it did in Tahoe (complete with the older Spotlight and Siri designs). Set up Golden Gate in a virtual machine on top of macOS, and it never offers Apple Intelligence or downloads those models. It works perfectly fine (accounting for the additional overhead that comes from virtualization, anyway). This is presumably how Macs in the EU will behave, since Apple still hasn’t committed to shipping the Apple Intelligence updates there yet. For those who have been following Windows 11’s development since it launched in late 2021, this may feel familiar. Microsoft has been aggressively pushing its AI features on users, and the company has a long history of failing to respect its users’ choices. It’s possible to turn off or uninstall many of the AI-powered features in modern Windows, including the nth iteration of the Copilot desktop app, the Windows Recall data scraper, and the inscrutably named “Click to Do.” But it’s not possible to turn off all of them, even if you chase down each of those individual toggles. I understand that Apple wants people to use its new technologies and that it thinks it finally has a version of Apple Intelligence that is worth including by default on all of its devices. And there’s some benefit to Apple and app developers in being able to assume that any Mac running Golden Gate has Apple Intelligence and its baseline AFM 3 Core model installed. Apple probably doesn’t want to have to maintain the old Spotlight and the new Siri-fied Spotlight in perpetuity. I could go on. But I still feel like users who want to opt out should still be allowed to—both because that toggle has already existed for two years and because there’s nothing keeping Golden Gate from functioning perfectly normally without Apple Intelligence installed. Some users don’t want AI at all; others might want to skip Apple’s entirely in favor of more capable third-party options. The way Golden Gate handles it, unless you’re in a country where Apple isn’t shipping this stuff, you have to set aside space for Apple Intelligence whether you want it or not. One change I do think is an improvement is the apparent retirement of the “soft” divider for window title bar areas. In Tahoe, the border between the title bar and the app’s contents existed, but it was invisible in some apps (Finder and Photos are two examples). An alternate “hard-style” divider was used in some other apps, retaining a bit of glassiness in the title bar but drawing a visible line between the title bar and your content and reducing the visibility of text and images located “beneath” the title bar. Sidebar icons regain the color they had in older releases, matching either the highlight color you’ve chosen in System Settings or a color of the app developer’s choosing in the default Multicolor mode. In Tahoe, these icons (and text labels) only showed color when actively selected. Actively selected sidebar items in Golden Gate get bolded text instead. But based on reactions I’ve seen during the betas, I expect these tweaks will be just enough to quell the loudest complaints, at least for people who were using Liquid Glass as an excuse not to upgrade. Apple likes to brag about the amount of Apple Intelligence processing that happens directly on-device without heading out to the cloud, and the company has made advancements here. But all the headlining features, including Siri, Image Playgrounds, and the Xcode coding assistants, are leaning heavily on cloud compute to work. AFM 3 Cloud is where you stop being able to run models on device and instead start relying on Apple’s own servers (running Apple’s own silicon) to handle your requests. And it’s the model that macOS will turn to for just about anything you ask the new Siri, or any requests for rewriting or summarizing information. The good news is that the AFM 3 Cloud’s improvements over the previous AFM Server model are much larger and less ambiguous than AFM 3 Core’s improvements over the old AFM Core. Apple Diffusion Model (ADM) 3 Cloud is also a cloud model running on Apple’s servers, and it powers the image generation features in Image Playground, Notes, and other places where you can generate or alter images. And then there’s AFM 3 Cloud Pro, which Apple says is for “agentic tool use and complex reasoning.” Apple had to turn to Nvidia GPUs running in Google’s cloud to make this model work, though the company says it is leveraging Nvidia, Intel, and Google technologies to enable privacy capabilities similar to the ones it offers for its own server hardware. Apple hasn’t said what limits users can expect, just that they will vary “by feature, request complexity, system demand, system policies, and other factors” and that additional usage “will be available for a fee.” Apple gates some Home app features, like video summaries from cameras, to upper-tier iCloud+ plans, and I wouldn’t be surprised to see other Apple Intelligence limits tied to iCloud upsells. By default, Siri uses the Apple models described above, primarily the AFM 3 Cloud model (you can tell the local models can’t do much here because most Siri queries fail if you switch off your Internet connection). ChatGPT is still available in System Settings as an extension, with an option to make it speak using a different voice than Siri’s. While Apple Intelligence as a whole can’t be turned off, users can opt out of the Siri AI beta at this point if they’d like. If you do this, you’ll be given the option to turn Siri off entirely or to use what Apple now dubs “Siri Classic.” In either mode, the UI for Spotlight reverts to its previous style. The new Siri experience is so much like Copilot or the AI answers in Google Search that it barely seems worth remarking upon, which is both a comment about how “behind” its rivals Apple is in the AI race and how “caught up” Apple feels now that it’s shipped something usable. One consistent barrier is that Siri can only occasionally help you with making your Mac do things; I’ve found that Siri can generally help change settings that are available in the Control Center, things like toggling Bluetooth or Light/Dark Mode, or turning screen brightness or volume up and down. Often it does that by presenting you with toggles you need to click yourself, but it can still help. It’s also capable of working with data inside individual apps—creating a note and dropping a few links in it, or pulling all of your week’s meetings from the Calendar app. But when it comes to manipulating more advanced system settings, Siri will look up and give you directions rather than trying to do anything for you directly. You could imagine a Siri AI that would help you set up an SMB file share, enroll/unenroll in an Apple beta program for updates, change your default browser, or even download and install apps. But that’s all well outside of the current Siri’s capabilities, as is directly manipulating files via the Finder. Even reading file metadata isn’t something Siri will do for you—asking it to show me the EXIF data for an image on my disk got me directions rather than displaying any of that information directly. Not that I think that’s a bad thing, necessarily. To give Siri carte blanche access to your filesystem and advanced system settings is to give a lot of users enough rope to hang themselves. That Siri can describe how to do many of these things is at least a step in the right direction that doesn’t put data at risk. App developers have a framework for integrating their own apps with visual intelligence, to make sure that the contents of their apps can be used while asking or answering questions. I find it annoying that this Ask Siri item appears at the top of nearly every single right-click menu almost anywhere in the operating system (apps that haven’t been updated for macOS 27 or to integrate with Apple Intelligence won’t show it). I’d also like there to be a toggle for hiding this icon that doesn’t require turning off Siri AI or reverting to the earlier version. The app’s UI changes to mirror the new Siri app’s. A simple sidebar on the left displays every image you’ve generated with the app so far, separated into stacks. Each stack represents a single prompt or modified image, with the original images and all their iterations stored in one stack so you can page through, delete, and re-iterate on them individually. Apple obviously has some kind of guardrails in place to try and prevent people from either depicting copyrighted characters or making images that are in the style of copyrighted characters. Some of the most fun I had with Image Playground was in gradually scooting around those prompts, one iteration at a time. Sometimes the app would see what I was doing and refuse to do anything else. I made a bald businessman look enough like Shrek from the movie Shrek that it would no longer let me add anything to the image, let alone change the background to look like a “swamp” or a “bog” or a “dirty lake.” But I could often iterate to an identifiable likeness of a copyrighted character in a way the app would refuse to do if I tried it all at once. The process of trying to defeat the app’s safeguards were actually kind of instructive. Generating an “italian plumber with a red hat” made a guy who had a giant mustache, even though I didn’t mention facial hair at all. Creating a “round pink creature” automatically made one with little flipper-like arm appendages, even though I didn’t say anything about arms. The data these models were trained on casts a long shadow and announces itself in subtle ways. Safari begins suggesting groups (or automatically grouping tabs) once you have at least three open tabs that pertain to the same loose topic (“iPhone 18 Pro,” “local AI”) or that come from the same site (Google Drive). For tabs where Safari has a group to suggest, users can click a down arrow icon on the right side of the tab that shows the suggested name for the tab group, the other tabs that will be added to the group, and a big button offering to let Safari start doing this automatically. RSS and RSS readers are both still things, and Ars is one of many sites that still offers an RSS feed from a page that looks like someone made it in 2007 and then forgot about it or that you have to find via search because the actual site doesn’t expose it anywhere. RSS was (and sometimes still can be) a reliable and algorithm-free way to get the latest updates from a curated selection of feeds. And while it doesn’t really have anything to do with RSS at all, it’s the first thing I thought of when I heard about Notify Me, Safari 27’s new feature for flagging updates to specific pages of users’ choosing. Navigate to a page that you’d like to be notified about—a news site or company blog you want to keep an eye on, for example, or a page for pre-ordering the next Star Trek Lego set or a specific MacBook configuration on Apple’s refurbished site. Click the little catch-all menu icon on the left side of Safari’s address bar and click Notify Me. Notify Me notifications show up just like any other Safari notification, with the browser’s icon, the name of the site, and a notice that a change was detected. Safari does not have to be running for Notify Me to work, though you need to allow the browser to send notifications in the first place, of course. Notifications sync via iCloud to other devices running iOS/iPadOS/macOS 27. To view and edit any Notify Me entries, you’ll find them under their own entry in the Websites tab of Safari’s settings. Apple lists several examples across four broad types of extensions to show users what is possible—things like “make the web page grayscale” or “show a colorful and explosive firework every time I click.” Type a description, and Apple will first show you a handful of Mac App Store extensions that might be able to do the job. If none are of interest, and if Describe an Extension can handle what you’re asking for, it will generate an extension within a couple of minutes. Iterating on an extension is fun, but there’s really no “history” or “undo” function for either reverting a change or trying several alternate suggestions. You just keep building on top of what you’ve made until you get what you want or until something errors out and you need to try again. And if you’re trying to recreate an extension, either for yourself or someone else, you just have to hope that (1) you can remember what you told the system to do and (2) that it interprets your inputs the same way and spits out the same result each time. Like the other Safari features here, these extensions can be synced between devices and between platforms using iCloud, though you’ll need to enable them and set up permissions when getting them to work on a new device. What you can’t do, possibly partly because Safari extensions can’t be distributed as standalone bits of software anymore, is export and share extensions with other users, or archive them for yourself. “For “real” developers, I expect the “Describe an Extension” feature will end up feeling too toy-like to use for serious work. But it is a legitimately cool/fun feature for anyone trying to close a specific Safari feature gap and who can do so without running up against the tool’s limitations. In this WWDC session video, Apple’s Jen Simmons reviews some other WebKit highlights for the year, noting that most of the team’s effort this year has been devoted to fixing bugs and paying down technical debt. The blog post for the initial Safari 27 beta calls out “525 fixes,” and that list has presumably gotten longer since June. That work extends to underlying web standards and, in a handful of cases, even fixing the standards bodies whose job it is to define and update the standards. Simmons calls out the revival of an SVG Working Group that Apple revived to help resolve issues in version 2 of the specification. This WWDC session video highlights the image quality improvements, particularly for intensely noisy or low-light photos. Apple says it’s running a new CoreML model on your device’s Neural Engine cores to enable the improvements, something that Golden Gate can always assume that your Mac has because it’s always running on an M1 or newer. If you have older RAW photos sitting around, it may be worth pulling them up in Golden Gate to see whether the quality improvements are noticeable. I’ve always liked the idea of the Shortcuts app, and I’ve made several simple ones that save me a bunch of time (one for making a PDF slide deck into a stack of PNGs, one for resizing huge images quickly, one for converting from PNG to JPG, one for automating connections to my home NAS, and so on). But for anything more complicated, I find myself getting frustrated stumbling through Shortcuts’ syntax and its trial-and-error, good-luck-figuring-out-what-went-wrong testing system. In theory, a feature that lets me tell Shortcuts what I want without having to figure out how to tell it what I want would make it much more useful. But as with Describe an Extension, I spent much of my time with Apple Intelligence in the Shortcuts app running up against its limitations. As Apple has built it, apps and services on your system offer short pre-set lists of actions to Shortcuts that you can then use to build a workflow. Apple Intelligence won’t try to color outside those lines to build a workflow that does other things. I used the new Shortcuts to update my NAS shortcut to work from anywhere—now it checks to see if my VPN is enabled, enables it if it turns off, and then connects, letting me use it from off-network. That worked well because it uses existing, well-defined actions in Shortcuts. An effort to create a Shortcut for automating the setup of a new Mac—downloading and installing apps from various sources, tweaking system settings, and cleaning up after itself when it was done—failed almost instantly. You can cobble something together for this, using commands for downloading and opening files from the Internet and maybe the odd Terminal command. Some other improvements that are worth calling out: Shortcuts can store their own data now, which can give them “memory” of their past actions. You can integrate Apple Intelligence into Shortcuts too, either by calling on cloud or on-device models. And Automations are a bit easier to work with because they can be built directly into Shortcuts rather than existing as a totally separate part of the app. But this is the last macOS release that will offer Rosetta 2 for general-purpose apps. Bits and pieces of the technology will hang around in macOS 28 and (possibly) future releases, but Apple has said that it will only work for older games that still have Intel-specific dependencies. Apple has started the process of warning users and developers that Rosetta support is ending, using a strategy similar to the one it used when it phased out support for 32-bit macOS apps a few years ago. Golden Gate still runs Rosetta 2 with no limitations, but it does display notifications warning that support for these apps is going away in the next major macOS release. In the General > About section of the System Settings app, Apple has added a new “Intel-based Apps” section that will show users all the recently launched Intel-exclusive apps on their system. Universal binaries don’t appear on this list—Apple has no issues with apps that choose to support Intel Macs in addition to Apple Silicon models—but the intent is clearly to push users into upgrading or moving away from older versions of apps that haven’t made the jump. Apple would probably prefer that you buy one of its own very expensive monitors to use with your Mac desktops and laptops, all else being equal. But Apple doesn’t offer monitors for everyone, particularly if you want cutting-edge display technologies, oddball resolutions, or screens with prices measured in hundreds of dollars rather than thousands. So macOS has to support other things. Golden Gate closes one gap in macOS’s display support by adding native support for ultrawide monitors with up to 5K resolutions and 120 Hz refresh rates; Apple hasn’t listed exactly which chips work with which resolutions and refresh rates, but “5K” in ultrawide-land usually refers to a 5120 x 1440 monitor. Even the basic M1 can support one of Apple’s 6K 6016 x 3384 displays at 60 Hz, so anything other than the MacBook Neo should run at at least 60 Hz on those 5K ultrawides. The Mac has supported HDR display output for years, but Golden Gate is the first release to actually extend HDR to the macOS user interface. And this looks quite nice if you happen to have monitors in your setup that support it or a MacBook Pro with an internal XDR display. Liquid Glass elements in HDR mode have a subtle-but-noticeable bit of extra definition and pop to them in Golden Gate, particularly in light mode, which further helps with readability on top of the other Liquid Glass adjustments Apple has made this year. First, Apple has added provisioning options for macOS guest VMs. Normally, when you start a new macOS VM using the Virtualization framework, you have to go through the Mac’s first-time setup flow just like you do when you’re setting up a new Mac or a fresh bare-metal macOS install. In Golden Gate, users can specify a username and password for a local user account, skipping the other setup screens entirely. You’ll be able to set that account to automatically log in if you’d like, and you can enable SSH access as well. These options make it a lot faster to fire up a macOS VM for testing, whether you’re running it on your desktop or on a headless system you’re using as a server. New additions to the vmnet framework now allow for more flexible network configurations, beyond the typical host mode/bridged mode options. Create a virtual network to allow VMs to communicate directly with one another or create multiple virtual networks to simulate the experience of communicating between separate physical networks. DHCP and port forwarding settings can also be configured. Another major improvement is Accessory Access, Apple’s name for a framework that allows USB passthrough to guest operating systems. Access to USB devices can be passed to running VMs through a menu bar icon, and once a device has been passed through to a guest operating system, that OS can work with it like it could if it were physically plugged in. Interact with the menu bar icon again to use the USB accessory with the host OS again, or pass it through a different VM. Developers of virtualization apps can choose to “indicate interest” in just a subset of USB devices, filtered by device class or subclass, by product ID, or by other characteristics. But I suspect that most apps will simply choose to let their users pass through any plugged-in USB device. Accessory Access explicitly only mentions USB accessories and not those attached via Thunderbolt or any other bus or interface. For attaching other kinds of hardware, guest VMs in Golden Gate support custom Virtio devices for Linux guests, for low-overhead communication with other kinds of hardware. (Apple says that Linux guests in Golden Gate can take advantage of EFI Secure Boot support.) Virtualization apps will need to be updated to take advantage of these new features rather than getting them for free. Using DiskImageKit images, you could create multiple VMs that all use the same read-only base layer—a fresh macOS install, for example. Then overlay layers record changes, so you could install different apps or sign in with different accounts in each VM you set up, while still retaining that fresh base image if you ever want to spin up another VM. Because the images for overlay layers are sparse, they only use enough of your physical disk space to record whatever changes you’ve made. Apart from supporting full-fat VMs, Apple has created a new container tool that works with Open Container Initiative (OCI) images, as an Apple-developed alternative to tools like Docker and Podman. (The container tool is compatible with macOS Tahoe as well, but not older versions of macOS.) While a virtual machine exists to convince the guest software that it is running on actual hardware—disk images to simulate a physical drive, CPU and RAM allotments, and so on—container images mainly just include individual apps, or apps bundled with the dependencies they need to run. You could run multiple apps that have database dependencies, for example, without making them all coexist in the same database. As an OCI-compatible tool, container uses commands for pulling, running, and working with images that you’re already used to if you’ve spent much time with the CLI for Docker or Podman. When you first run it, the tool will offer to download a Linux kernel for you if you don’t already have one—there still needs to be a base operating system under there somewhere. This isn’t a full-on Docker replacement, by any means. There’s no stand-in for the Docker Desktop GUI, and no replacement for Docker Compose that doesn’t rely on third-party tools. But for quickly pulling and running simple containers, it’s a nice addition. The Apple Filing Protocol (AFP) was the Mac’s primary networked file-sharing protocol starting all the way back in System 6 and remained the default until 2012’s OS X Mountain Lion release, which was the last version of the operating system to include any updates to the protocol. AFP was also the main file sharing protocol for the Time Capsule, Apple’s first-party router-and-NAS device that it sold from 2008 to 2018. Starting in 2013’s Mavericks, Apple switched to using SMB file-sharing by default, originally a Windows protocol but now widely used as the default on most Windows, macOS, and Linux-based computers and NAS devices. And AFP support has atrophied since then; 2020’s Big Sur release made it impossible to host an AFP file share on a Mac, and the protocol was declared deprecated in the macOS Sequoia 15.5 update. Golden Gate finally removes all AFP support from macOS, and with it the ability to back up to any Time Capsule devices that are still kicking. Many better router and NAS devices are available for both Wi-Fi networking and networked Time Machine backups now, but if you’d like to hold on to a Time Capsule for some reason, this TimeCapsuleSMB project promises to support modern SMB networking on the aging hardware. In another blow to anyone holding onto an outdated router, Apple’s release notes say it has deprecated the AirPort Utility, the app used for administering its routers and Time Capsules. This app’s design is stuck in an era of macOS so old that I’m not even sure which one it is—somewhere in the Mac OS X 10.5-to-10.8 range, I think—which along with “no longer making or selling routers” is a pretty clear sign of Apple’s disinterest. The release notes say that AirPort Utility “is no longer included with new clean installations of macOS,” though all of my Golden Gate Macs still have it. Your mileage may vary. Regardless of whether you have it or not, “functionality is not guaranteed starting in macOS 27,” and we’d bet out-and-out removals of the app will follow eventually. It’s been over 15 years since Apple last introduced a Mac with an optical drive built-in, but the operating system’s built-in DVD Player app and the DVDPlayback framework still kept soldiering on for anyone who connected an external drive. As of Golden Gate, their days are numbered. For the initial Golden Gate version, it’s just the framework that has been removed, and Apple says “apps cannot build against it.” The DVD Player is still present and functional. But the framework “will be removed in a future macOS release” and will presumably take the DVD Player app with it. Whether that future release is in the 27.x series or 28.x series, Apple didn’t say. As is our tradition, we have collected a bunch of minor changes that are worth documenting but don’t quite merit a section on their own. At this point, new desktop wallpapers are expected. What’s a little surprising to me is how long it’s been since Apple pruned its list of wallpapers. There are even stock wallpapers from the macOS 10.15 Catalina release still on the list here. There’s nothing wrong with that, but it’s sort of funny to me to still include callbacks to an OS version so old that it predates the M1. Some apps in Golden Gate will automatically offer to generate filenames for newly created files based on a summary of what’s in them. It’s not available in the initial release of Golden Gate, but “a future software update” is adding a feature to the Passwords app that will let it try to automatically sign into accounts with “weak or compromised passwords” and change them to stronger automatically-generated passwords. The current Passwords app already flags weak or reused passwords and passwords that have been exposed in known data breaches, but this takes those notifications a step further. Web login pages behave unpredictably enough that there will almost certainly be passwords that can’t be fixed automatically, but anyone who has set up a new password manager for someone else knows that the process of changing all the weak and reused passwords the first time can be pretty daunting. Automating any part of that could save a lot of time. Golden Gate (and iOS 27 and iPadOS 27) includes another Clarus Easter egg. The System Settings menu for adjusting Liquid Glass is scrollable, to give you a preview of how your chosen level of glassiness will look when overlaid against different kinds of images and text. If you scroll all the way to the bottom and keep scrolling, though, an unseen hand will type “Moof!” into the search bar, and a 3D-illustrated Clarus will appear. Cute! Howard Oakley at the Eclectic Light Company is probably the Internet’s best documenter of XProtect, essentially a quietly updated anti-malware service that Apple updates continuously with new definitions and other updates. In Golden Gate, XProtect “may now restrict access to app data that is commonly targeted by malicious software.” Security consultant Wojciech Reguła breaks down what this actually means in this blog post: Apple is protecting folders for certain apps using some of the same protections it applies to sandboxed apps from the Mac App Store. As of Golden Gate’s release date, the list of apps that merit this treatment include Discord, Google Chrome, Brave, Microsoft Edge, Firefox, Ledger Live, Exodus, and Wasabi. In most ways, the Golden Gate update is just what macOS needed after the Tahoe release. It refines (and, where necessary, tones down) the big visual redesign from last year. It adds a bunch of little nice-to-have things, like HDR in the macOS UI, Markdown support in Notes, the virtualization changes, and the better cleanup tools in Photos, that have quickly become part of my macOS workflow. And despite myself, I actually kind of enjoy Apple Intelligence this time around. Image Playground is fun now. Siri is significantly more useful for basic factual lookups, without nearly as many “I found some results on the web” bonks. The biggest downside is that if you still don’t want Apple Intelligence, and if you don’t live in the EU or a country where it’s not included with the operating system yet, the toggle for turning it off and ditching the many gigabytes’ worth of models it installs is no longer available. For most Mac users, though, Golden Gate is a no-brainer update with few major downsides. It won’t hurt anything to wait until the 27.1 or 27.2 release if you’re already on macOS 15 Sequoia or macOS 26 Tahoe and you’re not in a hurry. But I’ve found Golden Gate to be pretty stable as a daily driver from the very first developer beta, and finishing up the review you’re reading now is the last barrier to me installing it on the last of my Tahoe Macs.