For owners of older Intel Macs, the immediate effect is likely to arrive one app update at a time rather than as a single switch turned off by Apple. But the broader hardware boundary is now clear: Intel Macs remain supported by released macOS Tahoe 26.5.1, while forthcoming macOS 27 Golden Gate supports Apple-silicon Macs only. For developers, the App Store change can reduce the burden of maintaining a universal binary, but it also makes the supported-hardware decision more consequential.
The Mac App Store rule is real, but it has boundaries
Apple’s App Store Connect release notes say that a universal macOS app may drop support for Intel-based Macs if its minimum system version is macOS 13 or later. The practical instruction is simple: remove the x86_64 architecture, then upload and submit that build to the Mac App Store.
That makes an ARM64-only submission an available route for qualifying developers. An ARM64-only build is designed for Apple-silicon Macs rather than Intel-based Macs, while a universal build includes code for both processor families.
Three limits are central:
- The rule applies to universal macOS apps, not automatically to every Mac App Store title.
- The app must have a minimum deployment target of macOS 13 or later.
- It is a permission to change an app’s architecture, not an order that developers must abandon Intel Macs.
That final distinction matters to customers. A developer can decide that its Intel audience remains worth supporting and continue to provide a universal build. Conversely, an eligible developer can choose to concentrate future work on Apple silicon. The policy creates an option; it does not guarantee that every app will take it, or specify when any individual developer will do so.
June policy, September reminder
Apple publicly recorded the rule on June 18, 2026. That date is significant because reports of a September 1 developer email could otherwise leave the impression that Apple had only just introduced the capability.
The reported email appears consistent with Apple’s public rule: it said eligible universal Mac App Store apps could remove Intel support, emphasized development and download or on-device-size benefits, and directed developers toward ARM64-only builds. But the public release notes show the underlying permission was already in place months earlier.
The most accurate reading is therefore not that Apple suddenly changed the Mac App Store’s architecture rules in September. September appears to have brought a follow-up or targeted communication about a June-documented policy.
This correction is more than calendar trivia. It changes how developers and users should interpret the news. A June rule gives software makers time to plan architecture changes, whereas a supposedly abrupt September mandate would suggest an immediate platform cutoff. The evidence supports the former interpretation, not the latter.
What changed from the earlier ARM64-only path
It would also be wrong to say Mac App Store developers were previously required to include Intel support in every case. Apple had already permitted an app to support only Apple-silicon Macs under a different condition: the app needed to require at least macOS Monterey 12 and to have never supported Intel-based Macs.
The June policy is meaningful because it accommodates a different group—existing universal apps that meet the macOS 13-or-later deployment requirement and now want to remove their Intel architecture.
In plain terms, an all-new app that was Apple-silicon-only from the outset already had a route to remain that way. The newer rule is especially relevant to developers that once served both Intel and Apple-silicon customers but now want to end that dual-architecture commitment.
That is a more precise account than the broad claim that Apple has only now allowed ARM64-only Mac App Store software. It also explains why older applications could be affected differently from new ones: an app’s support history and its minimum macOS version are material to the submission decision.
Why developers may make the move
The reported September communication identifies several practical incentives: less development work and smaller downloads or on-device footprint. Those are intuitive consequences of no longer carrying a second processor architecture in a distributed app package.
For a developer, ending Intel support can also simplify the product decision in a broader sense. Instead of maintaining a single release across two Mac processor families, the developer can focus its Mac App Store build on Apple silicon. That does not prove every application will become better or faster; performance and feature outcomes depend on the individual software. But it reduces one explicit compatibility dimension.
The decision will not be equally attractive to all developers. Businesses with customers on eligible Intel hardware may see continuing value in universal releases. Smaller utilities, consumer apps, and products whose users have largely moved to Apple silicon may reach the opposite conclusion. A title’s update cadence, customer base, macOS requirement, and revenue model can all influence its choice.
There is also a customer-relations cost to consider. Removing Intel support may make the newest version unavailable to part of an established user base. Developers that take this route will need to communicate minimum requirements clearly, particularly where an update changes what Macs can run the current release.
The immediate reality for Intel Mac owners
The architecture policy does not mean Intel support has vanished from released macOS software. As of September 3, 2026, Apple identified macOS Tahoe 26.5.1 as the latest released macOS. Its compatibility information includes several Intel models, including the 2019 16-inch MacBook Pro, the 2020 four-port 13-inch MacBook Pro, the 2020 27-inch iMac, and the 2019 Mac Pro.
That gives Intel Mac owners an important present-tense reality check: supported Intel hardware still exists within the latest released macOS named by Apple. The newly flexible App Store submission policy should not be portrayed as proof that Apple has already ended all Intel Mac software support.
The next major macOS release establishes a firm hardware cutoff, however. Apple’s published macOS 27 Golden Gate compatibility list supports Apple-silicon Mac models only. Intel Mac hardware is not supported by macOS 27.
The likely near-term result is uneven application support rather than a synchronized cutoff before Golden Gate arrives:
- Some apps may continue to update on Intel Macs through universal releases.
- Some developers may leave an older Intel-compatible version in place while directing new development to Apple silicon.
- Other apps may release an ARM64-only update that an Intel Mac cannot run.
Independent reporting says users can continue using the last version compatible with their machine. That is useful guidance, but it is not a blanket assurance that every favourite app will remain available indefinitely. Continued usability can also depend on whether software is already installed, whether it remains on the App Store, whether a developer maintains it, and whether online services or associated components continue to operate.
The safest practical approach for an Intel Mac owner is to identify important applications now, check each developer’s stated system requirements before updating, and keep realistic expectations about long-term feature updates. A working installed version is not the same thing as an ongoing commitment to future compatibility.
Do not confuse Intel Macs with Intel-only apps
The coming Rosetta change affects a different group: owners of Apple-silicon Macs who still rely on apps compiled only for Intel processors.
Rosetta is Apple’s compatibility technology for running Intel-based applications on Apple-silicon Macs. It remains generally available on Apple-silicon hardware through macOS 27. Apple also says macOS 27 is the final macOS release with general Rosetta support for Intel-only apps.
Starting with macOS 28, Rosetta functionality is limited to certain older, unmaintained games that depend on Intel-based frameworks. That exception is narrow. It should not be read as a general promise that ordinary Intel-only productivity, business, utility, or creative applications will keep working after the transition.
This produces two distinct compatibility questions:
- Can an Intel-based Mac run a newer version of an app? The answer increasingly depends on that developer’s architecture and operating-system requirements. Intel Macs also cannot upgrade to macOS 27.
- Can an Apple-silicon Mac run an older Intel-only app? The answer depends on Rosetta and, beyond macOS 27, is subject to the sharply limited older-game exception.
These are easy to blur because both involve “Intel apps,” but they affect opposite hardware populations. An Intel Mac cannot benefit from Rosetta’s translation layer for Apple-silicon software. An M-series Mac, meanwhile, may run an Intel-only program today through Rosetta but should not assume that arrangement survives a future macOS upgrade.
For Windows users who also administer Macs, this distinction is especially practical. Inventory should record both the Mac’s processor family and whether each critical app is native Apple silicon, universal, or Intel-only. A list that records only “Mac-compatible” software will not be enough for upgrade planning.
macOS 27 has a confirmed hardware boundary, not a release date
Apple describes macOS 27 Golden Gate as coming in fall 2026, but it has not announced a specific public release date. Apple has confirmed a September 9 event, yet that event confirmation does not establish a Golden Gate launch date.
Users should therefore separate two facts. The hardware decision is published: macOS 27 supports Apple-silicon Macs and excludes Intel Mac hardware. The calendar decision remains unannounced: there is no confirmed public date for Golden Gate’s final release.
That distinction matters for practical planning. An Intel Mac owner can treat Tahoe 26.5.1 as the current supported release and Golden Gate as the next major-version endpoint for Intel hardware, without assuming that a September event is itself an operating-system release schedule.
A transition with a longer history than the headlines imply
Apple has changed Mac processor architecture before. It announced plans to move to Intel microprocessors on June 6, 2005, while saying Intel-based Macs would begin arriving in 2006. The announcement date should not be confused with the first Intel Mac shipments.
Apple then announced its transition to Apple silicon on June 22, 2020. At that point it presented Universal 2 binaries as the bridge: one application binary supporting both Apple silicon and Intel Macs.
Six years later, the company’s silicon line has reached M6, introduced on August 25, 2026. Calling that the sixth numbered base M-series generation is reasonable when counting M1 through M6, though it should not obscure the different variants within the broader family.
The current App Store change is best understood as a late-stage evolution of the Apple-silicon transition. Universal applications remain possible, but Apple now gives eligible developers a sanctioned route to stop including Intel code even when their apps previously supported both architectures.
What to do next
Developers with universal Mac App Store apps should first establish whether their minimum system version is macOS 13 or later. If it is, ARM64-only submission is now permitted. The next decision is commercial and technical rather than procedural: whether the benefit of a smaller, single-architecture build outweighs the value of continuing to serve Intel Mac customers.
Intel Mac owners should avoid assuming either immediate abandonment or indefinite continuity. Tahoe 26.5.1 still supports specified Intel models, but macOS 27 will not. Individual app roadmaps may also diverge as more developers exercise the option to drop Intel code. Check requirements before major updates and preserve access to essential installed software where appropriate.
Apple-silicon owners with old Intel-only applications face a separate deadline: Rosetta remains broadly available through macOS 27, while macOS 28 narrows it to a particular legacy-games exception. Before upgrading beyond macOS 27, identify critical Intel-only tools and seek native Apple-silicon or universal replacements.
The headline is not that Intel Macs stopped working in September, nor that every Mac app must become ARM64-only. It is that Apple has made it easier for a specific class of existing Mac App Store apps to leave Intel behind, while its published macOS 27 compatibility boundary makes the longer-term endpoint for Intel Mac hardware explicit.