I’m building an iOS app that uses FamilyControls + ManagedSettings + DeviceActivity. Goal: temporarily “unlock” a shielded app for N minutes, then automatically re-apply the shield when the timer expires. What I do: In the main app, when user picks an expiry (e.g. 15 min, 30 min). I start a non-repeating DeviceActivity schedule and remove the app’s ApplicationToken from ManagedSettingsStore().shield.applications. I also store activeUnlockBundleID etc. in an App Group so the DeviceActivityMonitor extension can re-lock at the end. Expected: DeviceActivityMonitor.intervalDidEnd(for:) is invoked when the non-repeating interval ends, and I re-add the token to the shield set. Actual: The app does not re-lock when the interval expires. I added OS logs as well as “debug local notifications” from the DeviceActivityMonitor extension in: init() intervalDidStart intervalDidEnd eventDidReachThreshold None of these logs or notifications ever appear, which suggests the extension is never invoked (or cannot schedule local notifications or OS logs). Environment: Device: iPhone 17 Pro iOS 26.3.1 Xcode 26.4 Running on a physical device Notification permissions for the app: granted App + extensions are in the same App Group entitlement. Extension Info.plist has: NSExtensionPointIdentifier = com.apple.deviceactivity.monitor NSExtensionPrincipalClass = $(PRODUCT_MODULE_NAME).DeviceActivityMonitorExtension Questions: Are there known limitations/requirements for DeviceActivityMonitor callbacks where intervalDidEnd doesn't to fire? Is posting local notifications / OS Logs from a DeviceActivityMonitor extension supported/reliable? If not, what’s the recommended way to verify the extension is invoked? If this looks like a platform bug, should I file Feedback Assistant? If so, what logs/artifacts are most useful?

Hi everyone, I’m the developer of one sec, an app used by a large number of users globally to reduce time spent on social media and to build healthier digital habits. Because of this, we rely heavily on Apple’s Screen Time / DeviceActivity / FamilyControls, ManagedSettings APIs – and unfortunately, we’re seeing increasingly severe issues in production that directly impact hundreds of thousands of real iOS users. During the past years, we have been busy filing dozens of feedback requests for different Screen Time issues – and there has been no response from Apple at all. Developer Relations might be able to "confirm" that the bugs are present and that they ended up with the right team – but they are never addressed, neither are workarounds provided. Instead, the situation gets worse and worse. iOS 26 introduced a series of heavy regressions (which have been reported via Apple’s official bug report tool "Feedback Assistant" on iOS 26 beta 1 in June 2025 – and have not been addressed 10 Months later). This is very frustrating for us as developers, but also for our end-users who run into these issues every day. In the end this impacts our ability to build an amazing product and hurts revenue (which affects both us and Apple). 1. DeviceActivity thresholds are not firing at all This affects both: our app’s usage of the API and Apple’s own Screen Time limits Radars: FB22304617, FB20526837, FB15491936, FB12195437, FB15663329, FB18198691, FB18289475, FB19827144 2. Screen Time usage data is clearly corrupted Websites showing hundreds of hours per week Up to ~20 hours per day of usage reported for a single domain Radars: FB22304617, FB17777429, FB18464235 3. DeviceActivity thresholds reaching threshold immediately Newly introduced with iOS 26 Reported on iOS 26 beta 1 in June No response so far / no workaround DeviceActivity calls didReachThreshold immediately after creating the DeviceActivityEvent – instead of waiting till the defined threshold is actually reached. Radars: FB13696022, FB18351583, FB21320644, FB18927456, FB18061981 4. Randomly Randomizing ApplicationTokens From time to time, and without consistency, Screen Time suddenly provides new, random, unknown tokens to my app in the ShieldConfigurationDataSource and ShieldActionDelegate. This has been reported on many times before here on the dev forms, many many years back already: https://forums.aninterestingwebsite.com/forums/thread/756440 https://forums.aninterestingwebsite.com/forums/thread/758325 https://forums.aninterestingwebsite.com/forums/thread/758325?answerId=793267022#793267022 Radars: FB14082790 and FB18764644 5. Moving Tokens from one ManagedSettingsStore to Another Removing an ApplicationToken from one SettingsStore and then adding it to another while the target app remains in foreground leads to the re-use of the ShieldConfiguration. Which can be wrong in many scenarios. It is not possible to request a re-request of the ShieldConfiguration in that scenario. Radar: FB14237883 6. Unable to Open Parent App (one sec) from Shield Many times, when a target app is blocked by a shield, the user wants to perform some action (e.g. to unlock more time for the target app via an intervention). That means, that somehow I have to forward the user from a ShieldActionDelegate back into my target app. Unfortunately, there’s no API for that. Many apps on the App Store rely on private API to achieve that, but that’s too risky for a popular app like one sec. Radar: FB15079668 7. Unable to Open Target App from an ApplicationToken When a user has completed an intervention within one sec, and they indend to to continue to the target app, there is no way that one sec can open the target app just from the token alone. Sure, there are URL schemes, but that means the user has to manually assign URL schemes to each ApplicationToken. That is not a very user friendly process (and in many cases impossible, because not every app registers URL schemes). It would be better if there was a way that my app could open a target app directly from an ApplicationToken, e.g. via an AppIntent that can be run on a button press. This way, the selected apps would remain fully private while still offering advanced functionality: struct OpenTargetAppIntent: AppIntent, OpenAppFromApplicationTokenIntent { func perform() { return .result(openAppFromApplicationToken: applicationToken) } } Radar: FB15500695 Summary Thanks a lot for taking the time to read my feedback. If you have any questions, please feel free to reach out to me any time. I’m always happy to provide more details, logs, and steps to reproduce in my radars / feedback requests or in-person in Cupertino. It would be extremely helpful if someone from the Screen Time / DeviceActivity engineering team could: Take a look at the listed radars. Work on bug fixes and be transparent about when fixes will be shipped. Provide workarounds in the meantime. We genuinely want to build great, reliable experiences on top of Screen Time – but in its current state, it’s becoming very difficult to depend on. – Frederik

[FB22327481] We are observing a rare issue where alarms scheduled using AlarmKit occasionally fire exactly at 12:00 AM, even though the alarm was scheduled for a different time. This issue happens only for a very small number of users (for example, 1–2 users per several thousand per day), but multiple reports confirm that the alarm goes off exactly at midnight. We also found that other developers are experiencing the same issue: https://aninterestingwebsite.com/forums/thread/815714 Observed behavior For a small number of users: Alarm fires at exactly 00:00 (midnight) This happens even when the scheduled alarm time is something like 7:30 AM Happens rarely Hard to reproduce internally Appears to happen only on real devices in production Additional notes We are using AlarmKit fixed schedules (not relative) Dates passed to AlarmKit are correct at scheduling time We do not intentionally create midnight alarms Issue seems random Question Are there any known limitations or edge cases with AlarmKit fixed schedules that could cause alarms to fire at midnight? For example: Invalid date fallback? Schedule expiration fallback? Time zone changes? Background refresh timing? Maximum schedule window? Any guidance would be appreciated, as this is affecting real users but is difficult to reproduce.

On iOS 26.3.1, deviceactivityd produces zero log output and never wakes a DeviceActivityMonitor extension at the scheduled time, even when the schedule is successfully registered. The extension's intervalDidStart(for:) method is never called. Steps to Reproduce: Create an app with a DeviceActivityMonitor extension subclass Add the com.apple.developer.family-controls entitlement to both the main app and extension targets Add a shared App Group entitlement to both targets In the main app, call: let center = DeviceActivityCenter() center.stopMonitoring([.myActivity]) let schedule = DeviceActivitySchedule( intervalStart: DateComponents(hour: 17, minute: 30), intervalEnd: DateComponents(hour: 17, minute: 45), repeats: true ) try center.startMonitoring(.myActivity, during: schedule) Verify registration succeeds: DeviceActivityCenter().activities.contains(.myActivity) returns true Background the app completely and wait for the intervalStart time (5:30 PM in this example) Expected Result: intervalDidStart(for:) is called in the extension process at 5:30 PM. ManagedSettings shields are applied without any user interaction. Actual Result: intervalDidStart(for:) is never called. No extension process is launched. Filtering Console.app for deviceactivityd on the device returns 0 messages — the daemon produces no logs whatsoever despite an active registered schedule. Shields are never applied. Additional Evidence: Third-party apps using DeviceActivity (e.g. Prayer Lock) successfully apply shields at scheduled times on the same device and OS version, confirming the framework is not universally broken Filtering Console.app for the app's bundle ID at the trigger time shows the main app process applying shields via a UNCalendarNotificationTrigger fallback — but no extension process activity DeviceActivityCenter().activities correctly lists the registered activity before and after the trigger time, confirming the schedule registration API is functional Environment: iOS 26.3.1 Physical device (required for FamilyControls) Development-signed build with com.apple.developer.family-controls entitlement provisioned App Group shared between main app and extension targets

I’m running into a hardware reality. MEMS sensor thermal drift. If a user zeroes out the tilt indoors at 20°C and then takes the phone outside in the cold, the accelerometer baseline shifts just enough as the device cools to throw off the readings. I want to apply a simple thermal compensation curve to the CoreMotion data to keep the "zero" perfectly level regardless of the weather. However, ProcessInfo.thermalState only gives broad buckets (nominal, fair, etc.) which doesn't help me calculate a continuous offset for a phone cooling down degree by degree. Is there any public API, or even a proxy metric, that can give me a rough battery or internal temperature integer? I don’t need high resolution decimals. Just a general device temp to offset the hardware drift. Any undocumented tricks or proxy metrics anyone has used to handle this?

Hello, I submitted a request for the Family Controls (Distribution) entitlement, but haven't received status update regarding approval/rejection etc. I submitted a previous contact support ticket as well. I'm wondering the timeline and also if my request went through - currently it says 'submitted' but it's remained this way for a while... I've had other developers in communities saying they were approved earlier, so curious if it's an app issue. Thank you

Hello, I am the developer of an app called one sec which helps users to spend less time on social media: https://one-sec.app Therefore, we make heavy use of the Screen Time API, and thus ManagedSettings and ShieldActionDelegate. One feature of one sec is the so-called “Doom Scroll Emergency Brake”. This blocks a target app after a certain usage threshold (e.g. 5 minutes) and requires going through an intervention (e.g. breathing exercise) to unlock more time. That added friction makes it very effective in reducing time spent on apps. One thing that is confusing for our users is the way they are prompted to unlock more time, if they want to. They have to: Have Push Notifications enabled for one sec Exempt one sec’s notifications from being delayed by AI prioritization (otherwise they are delayed by ca. 10s) Ensure that push notifications can be delivered during foci. Understand that they have to tap on the notification, which is not very straight-forward because it does not make sense from the user’s UX perspective. This is an artificial limitation of Apple’s screen time framework which has no reason (no security / privacy implications here…). Screenshots of the current flow attached. If would be much more reasonable if there was a new ShieldActionResponse.openParentApp value that can be returned from the completion handler of the ShieldActionDelegate.handle(…) callback. We have seen different apps use private API to achieve this, but we are afraid to do the same to avoid getting banned from the App Store. It would be fair if Apple would level the playground for all apps and offer such an API officially. – Frederik PS: Tracked under FB22347946, FB18846650, FB15500681, FB15079668, FB10393561 (all without responses so far…)

We are developing an MDM agent app that uses FamilyControls with .individual authorization to enforce Screen Time restrictions (app blocking, domain blocking via ManagedSettingsStore and DeviceActivityCenter). The Problem We are actively subscribing to AuthorizationCenter.shared.$authorizationStatus to detect authorization changes. However, when the user revokes the app's FamilyControls authorization through Settings (either via Settings > Screen Time > Apps With Screen Time Access, or Settings > Apps > [Our App]), the publisher does not emit any value. All ManagedSettingsStore restrictions are lifted immediately by the system, but our app receives no notification of this change. The only scenario where the publisher reliably emits is when a debugger is attached (i.e., running directly from Xcode). Without the debugger, the publisher is completely silent — even when the app returns to foreground. Code Example We tried subscribing directly to AuthorizationCenter.shared.$authorizationStatus with no intermediary, exactly as shown in the documentation: AuthorizationCenter.shared.$authorizationStatus .sink { status in print("[DIRECT] authorizationStatus emitted: \(status)") } .store(in: &cancellables) This subscription is set up at app launch and stored in cancellables. The result is the same — the publisher does not emit when the user revokes authorization in Settings without a debugger attached. Documentation Reference The documentation for authorizationStatus states: "The status may change due to external events, such as a child graduating to an adult account, or a parent or guardian changing the status in Settings." And: "The system sets this property only after a call to requestAuthorization(for:) succeeds. It then updates the property until a call to revokeAuthorization(completionHandler:) succeeds or your app exits." This suggests the publisher should emit when the status is changed via Settings, but in our testing it does not — unless a debugger is attached. What We Verified We tested with a development-signed build (which includes the com.apple.developer.family-controls entitlement), launched from Xcode, then disconnected the debugger, killed the app, and relaunched from the home screen. Scenario Publisher emits on revocation? Running from Xcode (debugger attached) Yes, immediately Development-signed build (no debugger) No — silent even on foreground return We also confirmed: MDM configuration profiles can disable Screen Time entirely, but cannot restrict the per-app authorization toggle — the user can always freely revoke the app's Screen Time access The Security Gap This creates a significant gap for parental controls use cases: User leaves the app (app goes to background) User goes to Settings and disables Screen Time access for the app All restrictions are immediately lifted User uses the device freely User re-enables Screen Time access and opens the app Everything syncs back to normal — administrator never knows Questions Is there any supported mechanism to receive a notification (background or foreground) when FamilyControls individual authorization is revoked? We are subscribing to AuthorizationCenter.shared.$authorizationStatus but it does not emit. Is the $authorizationStatus publisher expected to work only when a debugger is attached? Is this a known limitation or a bug? Can DeviceActivityMonitor extension detect authorization revocation? Based on documentation it appears limited to schedule/threshold events, but we haven't confirmed this. Is there a planned API improvement to address this gap? Environment iOS 26.2 Xcode 26.3 Swift 6.2.4 FamilyControls .individual authorization Related Threads Screen time API can be disabled easily Changing Screen Time Passcode does not protect apps

Is there a way to trigger the start of my app (just for a short time) when the iPhone is connected to CarPlay (and/or to a car via bluetooth). I could not find anything. I don´t want to apply for CarPlay as I have no intention to show something on the cars screen. Any even small tipp is welcome

I'd like to create a small helper app for new students do read/write User default settings. Since it was not possible using the UserDefaults class I decided to use the "/usr/bin/defaults". Unfortuntely it seems not to return anything. Debug output shows "Got data: 0 bytes" Here is a sample code: import SwiftUI func readDefaults(domain : String, key :String) -> String { let cmdPath = "/usr/bin/defaults" //let cmdPath = "/bin/ls" let cmd = Process() let pipe = Pipe() cmd.standardOutput = pipe cmd.standardError = pipe cmd.executableURL = URL(fileURLWithPath: cmdPath, isDirectory: false, relativeTo: nil) cmd.arguments = ["read", domain, key] //cmd.arguments = ["/", "/Library"] print("Shell command: \(cmdPath) \(cmd.arguments?.joined(separator: " ") ?? "")") var d : Data? do { try cmd.run() d = pipe.fileHandleForReading.readDataToEndOfFile() cmd.waitUntilExit() } catch let e as NSError { return "ERROR \(e.code): \(e.localizedDescription)" } catch { return "ERROR: call failed!" } // get pipe output and write is to stdout guard let d else { return "ERROR: Can't get pipe output from command!" } print("Got data: \(d)") if let s = String(data: d, encoding: String.Encoding.utf8) { print("Got result: \(s)") return s } else { return "ERROR: No output from pipe." } } struct ContentView: View { let foo = readDefaults(domain: "com.apple.Finder", key: "ShowHardDrivesOnDesktop") var body: some View { VStack { Text("ShowHardDrivesOnDesktop: \(foo.description)") } .padding() } } #Preview { ContentView() } This code works well e.g. for "ls" when the comments are changed for cmdPath and cmd.arguments. What do I miss in order to get it working with defaults?

Hi, Regarding the Brazil Digital ECA (DECA) requirements, which became effective on March 17, 2026. Following the guidance for regulated regions, we have been testing the eligibility check with the iOS 26.4 release. We previously confirmed that isEligibleForAgeFeatures was returning true for users in the Brazil region, which allowed us to verify our age-gating implementation. A few questions follow on this eligibility check: We have observed in manual testing that specific devices which returned true as recently as March 23rd are now returning false today, despite no changes to the OS build or account settings. Does this indicate a change in the server-side eligibility heuristics? Why would isEligibleForAgeFeatures stop returning true for a region where the law is now in force? Has the guidance on how to evaluate these properties for Brazil changed with the transition to the stable 26.4 release? Thank you!

Hi, we are developing a screen time management app. The app locks the device after it was used for specified amount of time. After updating to iOS 26.2, we noticed a huge issue: the events started to fire (reach the threshold) in the DeviceActivityMonitorExtension prematurely, almost immediately after scheduling. The only solution we've found is to delete the app and reboot the device, but the effect is not lasting long and this does not always help. Before updating to iOS 26, events also used to sometimes fire prematurely, but rescheduling the event often helped. Now the rescheduling happens almost every second and the events keep reaching the threshold prematurely. Can you suggest any workarounds for this issue?

Issue: after going through configuration steps for app clips, when I scan my QR code, my app clip does not appear, instead safari attempts to open the url as a web page. note: my aasa endpoint is never even getting called when scanning the QR code. Setup: App uninstalled in accordance with Apple Documentation "Users don’t install App Clips, and App Clips don’t appear on the Home Screen. Similarly, testers don’t install the beta version of your App Clip" testflight installed in accordance with Apple Documentation My app's Build 1.51.9 (1) uploaded and greenlit in testflight. My apple email is added as an internal tester is the same as my Apple ID for the device used. I have provided an aasa for the path: .well-known/apple-app-site-association. Here is my full url: https://akin-server-side-staging.onrender.com/.well-known/apple-app-site-association. { "appclips": { "apps": [ "8PJ28P9ZZ8.com.ElevatedUnderdogs.akin1.Clip" ] }, "applinks": { "details": [ { "components": [ { "/": "/appClips/referral/venueToUser" } ], "appIDs": [ "8PJ28P9ZZ8.com.ElevatedUnderdogs.akin1" ] } ] } } Here are my entitlements for my parent target: <?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd"> <plist version="1.0"> <dict> <key>aps-environment</key> <string>development</string> <key>com.apple.developer.applesignin</key> <array> <string>Default</string> </array> <key>com.apple.developer.associated-domains</key> <array> <string>applinks:akin-server-side.onrender.com</string> <string>applinks:akin-server-side-staging.onrender.com</string> <string>appclips:akin-server-side-staging.onrender.com</string> <string>appclips:akin-server-side.onrender.com</string> </array> <key>com.apple.security.application-groups</key> <array> <string>group.com.ElevatedUnderdogs.akin1</string> </array> </dict> </plist> Here are the entitlements for my app clip target: <?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd"> <plist version="1.0"> <dict> <key>com.apple.developer.associated-domains</key> <array> <string>applinks:akin-server-side-staging.onrender.com</string> <string>applinks:akin-server-side.onrender.com</string> </array> <key>com.apple.developer.parent-application-identifiers</key> <array> <string>$(AppIdentifierPrefix)com.ElevatedUnderdogs.akin1</string> </array> <key>com.apple.security.application-groups</key> <array> <string>group.com.ElevatedUnderdogs.akin1</string> </array> </dict> </plist> On App Store Connect in the Testflight section for this project and build: Build 1.51.9 (1), Test Information -> App Clip Invocations I have the following: copy pasted for convenience: "no variables":"https://akin-server-side-staging.onrender.com/appClips/referral/venueToUser", "Jeff referral":"https://akin-server-side-staging.onrender.com/appClips/referral/venueToUser?venueID=ChIJVaPxJnCej4ARyxiB9Tt2tG8&referrerName=Jeff" Here is the QR code I attempted to scan, https://akin-server-side-staging.onrender.com/appClips/referral/venueToUser?venueID=ChIJVaPxJnCej4ARyxiB9Tt2tG8&referrerName=Jeff

The Numbers.app reopens the last edited document when the app launches. If the document was moved to another folder in the Files.app while the app was not running, Numbers.app correctly tracks the file and reopens it. However, if the document was deleted in the Files.app and moved to Recently Deleted, Numbers.app does not reopen the document when the app launches. Question : How does Numbers.app detect that a document has been moved to Recently Deleted? Can third-party apps implement the same behavior? What I tested : If a file is moved while the app is not running, resolving a bookmark successfully tracks the moved file. Files that are deleted via the Files.app appear in Recently Deleted, but those files are actually moved to the following directories: iCloud Drive /var/mobile/Library/Mobile Documents/.Trash/ On My iPad /var/mobile/Containers/Shared/AppGroup/{UUID}/File Provider Storage/.Trash/ App sandbox Documents directory ([On My iPad]/[Any App]) /var/mobile/Containers/Data/Application/{UUID}/Documents/.Trash/ When resolving the bookmark after deletion, the bookmark still resolves successfully and returns the new file URL inside the .Trash directory. I tried the following checks on the resolved URL: Checking file existence Checking read/write accessibility Inspecting bookmark resolution results Using APIs related to NSTrashDirectory See https://aninterestingwebsite.com/forums/thread/813329#813329021 All of these behaved the same as when the file was moved to a normal directory. None of these checks allowed me to detect that the file had been deleted. Additional experiment: I suspected that the app might simply check whether the path contains ".Trash", so I performed the following experiment. If a .numbers file is moved to /var/mobile/Containers/Data/Application/{UUID}/Documents/.Trash/ then The file appears in Recently Deleted in Files.app Numbers.app does not reopen the document when the app launches However, if the same file is moved to Documents/Trash Documents/.Trashed Documents/Any Folder/.Trash then The file does not appear in Recently Deleted in Files Numbers does reopen the document when launched This suggests that Numbers.app is not simply checking whether the path contains ".Trash".

Hello, I’m honestly trying to understand what’s going on with the Family Controls API review process. I submitted my entitlement request on March 5, and as of today (March 19), there has been zero response. I also opened a support ticket on March 16, and that hasn’t received any response either. What’s confusing is that updates to another app on the same developer account were approved within days so clearly the account is in good standing and active. At this point, it feels like there’s no visibility into what’s happening: Is the request under review? Is it waiting for additional information? Or is it just sitting in a queue indefinitely? I understand that Family Controls is a sensitive API, but a two-week silence with no status or communication makes it very difficult to plan or ship features. For those who’ve gone through this: How long did your approval actually take? Did you receive any communication during the process? Is there any reliable way to get visibility or speed this up? Right now, this is blocking a feature we’ve invested significant time building, and the lack of feedback is the most frustrating part. Would really appreciate any insights.

Hi community. It's my first time trying WatchConnectivity kit. When I attempt to send ApplicationContext from phone, my debug print indicates that updateApplicationContext is working correctly. However, I'm having trouble receiving it form the paired watch simulator as nothing is shown on that end. Here are the code for the Phone Extension:     func sendDataToWatch(data: String) {         state.append(data)         do {             NSLog("sent by phone")             try WCSession.default.updateApplicationContext(["currentstate": state])         }         catch {             print(error)         }     } Here are the code for watch: func session(_ session: WCSession, didReceiveApplicationContext applicationContext: [String: Any]) {         print("Hello")         if let getState = applicationContext["currentstate"] as? [String]{             print("\(getState)")             self.state = getState[0]         }     } Any suggestion would be appreciated!

https://aninterestingwebsite.com/documentation/DeclaredAgeRange/AgeRangeService/isEligibleForAgeFeatures returns a bool. I assume that means that it will return True for the states where their laws are in effect. The TX law and the UT/LA/AZ laws have different requirements though: TX requires the app verify the user's age on every app launch. These other states require the app verify the user's age "no more than once during each 12-month period" A future law (Brazil maybe?) might do something else. How can we determine if the user is eligible for the TX versus other state requirements?

A screen time app I'm making has started telling users that their limit was reached even when they're far below their limit for the day (sometimes even at 0 minutes for the day). This issue only started happening after upgrading my software to iOS 26.2. Is this happening to anyone else? If so how have you found any solutions or does anyone know of any changes that could be causing this? Any help would be appreciated.

Hello, my app is frequently loosing / forgetting the Screen Time Permission that had been granted previously on iOS 26. I have experienced it myself, sysdiagnose is in this radar: FB18997699 But also also my App Store users who have updated to iOS 26 already have reported this bug. It would be great if Apple could ensure that this bug is addressed before iOS 26 is released to the public.

I am getting bug reports from users that occasionally the AlarmKit alarms scheduled by my app are going off exactly at midnight. In my app, users can set recurring alarms for sunrise/sunset etc. I implement this as fixed schedule alarms over the next 2-3 days with correct dates pre-computed at schedule time. I have a background task which is scheduled to run at noon every day to update the alarms for the next 2-3 days. Are there any limitations to the fixed schedule which might be causing this unintended behavior of going off at midnight?