I work for Brave, a browser with ~80M users. We want to introduce a new system for automatic updates called Omaha 4 (O4). It's the same system that powers automatic updates in Chrome. O4 runs as a separate application on users' systems. For Chrome, this works as follows: An app called GoogleUpdater.app regularly checks for updates in the background. When a new version is found, then GoogleUpdater.app installs it into Chrome's installation directory /Applications/Google Chrome.app. But consider what this means: A separate application, GoogleUpdater.app, is able to modify Google Chrome.app. This is especially surprising because, for example, the built-in Terminal.app is not able to modify Google Chrome.app. Here's how you can check this for yourself: (Re-)install Chrome with its DMG installer. Run the following command in Terminal: mkdir /Applications/Google\ Chrome.app/test. This works. Undo the command: rm -rf /Applications/Google\ Chrome.app/test Start Chrome and close it again. mkdir /Applications/Google\ Chrome.app/test now fails with "Operation not permitted". (These steps assume that Terminal does not have Full Disk Access and System Integrity Protection is enabled.) In other words, once Chrome was started at least once, another application (Terminal in this case) is no longer allowed to modify it. But at the same time, GoogleUpdater.app is able to modify Chrome. It regularly applies updates to the browser. For each update, this process begins with an mkdir call similarly to the one shown above. How is this possible? What is it in macOS that lets GoogleUpdater.app modify Chrome, but not another app such as Terminal? Note that Terminal is not sandboxed. I've checked that it's not related to codesigning or notarization issues. In our case, the main application (Brave) and the updater (BraveUpdater) are signed and notarized with the same certificate and have equivalent requirements, entitlements and provisioning profiles as Chrome and GoogleUpdater. The error that shows up in the Console for the disallowed mkdir call is: kernel (Sandbox) System Policy: mkdir(8917) deny(1) file-write-create /Applications/Google Chrome.app/foo (It's a similar error when BraveUpdater tries to install a new version into /Applications/Brave Browser.app.) The error goes away when I disable System Integrity Protection. But of course, we cannot ask users to do that. Any help would be greatly appreciated.

Hi Team, How can we fetch the macOS password requirement(for setting a new password) that are inforce during login for users? Is there a way to get this info in swift programming?

Trying to validate external reference identifiers with SecTrustEvaluateWithError Method by setting reference Ids to SecPolicyCreateSSL() & SecPolicyCreateWithProperties() But two concerns are - Validates for correct reference IDs but gives error for combination of wrong & correct reference Ids 398 days validity works mandatorily before reference Ids check. Is there any other to validate external reference Ids?, which give flexibility To pass multiple combinations of reference IDs string (wrong, correct, IP, DNS) To validate reference ID without days validity of 398. Please suggest. Any help here is highly appreciated.

Hello -- I am developing an Authentication Plug-in for the purpose of invoking login with no user interaction (headless). There seems to be sufficient documentation and sample code on how to implement a plug-in and mechanism, and debug the same, which is great. What I am trying to understand is exactly how to modify the login right (system.login.console) in order to accomplish my goal. Question 1: I had the idea of installing my mechanism as the first mechanism of the login right, and when invoked to set the username and password into the engine’s context, in the belief that this would negate the system from needing to display the login screen. I didn’t modify or remove any other mechanisms. This did not work, in the sense that the login screen was still shown. Should this work in theory? Question 2: I then tried modifying the login right to remove anything that interacted with the user, leaving only the following: <array> <string>builtin:prelogin</string> <string>builtin:login-begin</string> <string>builtin:forward-login,privileged</string> <string>builtin:auto-login,privileged</string> <string>MyAuthPlugin:customauth,privileged</string> <string>PKINITMechanism:auth,privileged</string> <string>builtin:login-success</string> <string>HomeDirMechanism:login,privileged</string> <string>HomeDirMechanism:status</string> <string>MCXMechanism:login</string> <string>CryptoTokenKit:login</string> </array> The mechanisms I removed were: <string>builtin:policy-banner</string> <string>loginwindow:login</string> <string>builtin:reset-password,privileged</string> <string>loginwindow:FDESupport,privileged</string> <string>builtin:authenticate,privileged</string> <string>loginwindow:success</string> <string>loginwindow:done</string> In place of builtin:authenticate I supplied my own mechanism to verify the user’s password using OD and then set the username and password in the context. This attempt appears to have failed quite badly, as authd reported an error almost immediately (I believe it was related to the AuthEngine failing to init). There’s very little information to go on as to what each of these mechanisms do, and which are required, etc. Am I on the wrong track in attempting this? What would be the correct approach?

Problem Statement: Pre-requisite is to generate a PKCS#12 file using openssl 3.x or above. Note: I have created a sample cert, but unable to upload it to this thread. Let me know if there is a different way I can upload. When trying to import a p12 certificate (generated using openssl 3.x) using SecPKCS12Import on MacOS (tried on Ventura, Sonoma, Sequoia). It is failing with the error code: -25264 and error message: MAC verification failed during PKCS12 import (wrong password?). I have tried importing in multiple ways through, Security Framework API (SecPKCS12Import) CLI (security import &lt;cert_name&gt; -k ~/Library/Keychains/login.keychain -P "&lt;password&gt;”) Drag and drop in to the Keychain Application All of them fail to import the p12 cert. RCA: The issues seems to be due to the difference in the MAC algorithm. The MAC algorithm used in the modern certs (by OpenSSL3 is SHA-256) which is not supported by the APPLE’s Security Framework. The keychain seems to be expecting the MAC algorithm to be SHA-1. Workaround: The current workaround is to convert the modern p12 cert to a legacy format (using openssl legacy provider which uses openssl 1.1.x consisting of insecure algorithms) which the SecPKCS12Import API understands. I have created a sample code using references from another similar thread (https://developer.apple.com/forums/thread/723242) from 2023. The steps to compile and execute the sample is mentioned in the same file. PFA the sample code by the name “pkcs12_modern_to_legacy_converter.cpp”. Also PFA a sample certificate which will help reproduce the issue by the name “modern_certificate.p12” whose password is “export”. Questions: Is there a fix on this issue? If yes, pls guide me through it; else, is it expected to be fixed in the future releases? Is there a different way to import the p12 cert which is resistant to the issue? This issue also poses a security concerns on using outdated cryptographic algorithms. Kindly share your thoughts. pkcs12_modern_to_legacy_converter.cpp

I am working on improving Keychain item storage secured with Face ID using SecAccessControlCreateWithFlags. The implementation uses the .biometryAny flag as shown below: SecAccessControlCreateWithFlags( kCFAllocatorDefault, kSecAttrAccessibleWhenUnlockedThisDeviceOnly, .biometryAny, &amp;error ) While this approach generally works as expected, I encountered a specific edge case during testing. On iOS 18.3.1 with Xcode 15.4, the following sequence causes the Keychain item to become inaccessible: Navigate to Settings &gt; Face ID &amp; Passcode and select Reset Face ID. Before setting up a new Face ID, tap the Back button to exit the setup process. Reopen the Face ID setup and complete the enrollment. Return to the app—previously stored Keychain items protected by .biometryAny are no longer available. This behavior appears to be a change introduced in recent iOS versions. In versions prior to iOS 15, resetting or deleting Face ID entries did not invalidate existing Keychain items protected by .biometryAny. This difference in behavior between iOS versions raises questions about the changes to biometric protection handling. Any suggestions are welcomed that might shine a light on what the best practice to use keychain access control and prevent the data to become unavailable.

Hello! We have code that extracts macOS Installer package (.pkg, .mpkg) signature information using APIs defined in <xar/xar.h> The code opens the package using ‘xar_open’ API like this. func open(file: String) throws(XarError) { xarfile = xar_open(file, READ) if xarfile == nil { throw .fileOpenError } } This code produces a clang warning in our CI build system when built for macOS 12 and up. 'xar_open' was deprecated in macOS 12.0: xar is a deprecated file format and should not be used. Question #1: What is the appropriate / more preferred way to extract signature information from an Installer package given that xar related APIs are deprecated? We use xar APIs to validate the package signature prior to installation to prevent packagers not signed by our team ID from being installed. Question #2: “xar is a deprecated file format and should not be used.”. Does this phrase refer to the file format that should be avoided or the API that extract signature information? We distribute our product using Developer ID method that using pkg/mpkg formats which I believe internally follow the same structure as xar files. I hope this message does not mean we should rethink the distribution method for our products. Thank you. Filed FB FB17148233 as well.

We recently transferred two applications to a different account, both of which utilize Keychain and shared app containers. Before transferring the first application, we anticipated losing access to the Keychain and took proactive measures by backing up data to the app’s private container in the final release prior to the transfer. During the app transfer process, we removed the shared container group ID from the old account and recreated it under the new account. In our testing, Keychain restoration from the local backup was successful, and users experienced no disruptions. However, after releasing the application, we observed that approximately 25% of our users not only lost their Keychain data as expected but also their shared app container data. As we have been unable to reproduce this issue internally, we are seeking your guidance on how to prevent a similar situation when transferring our second application. At this stage, we have not yet released any updates from the new account, and the Keychain data remains backed up in the app’s private container. We would appreciate any insights or recommendations you can provide to ensure a smooth transition for our users and make sure we can keep the data in shared container.

I'm using Secure Enclave to generate and use a private key like this: let access = SecAccessControlCreateWithFlags(nil, kSecAttrAccessibleWhenUnlockedThisDeviceOnly, [.privateKeyUsage, .biometryAny], nil) let attributes: [String: Any] = [ kSecAttrKeyType as String: kSecAttrKeyTypeECSECPrimeRandom, kSecAttrKeySizeInBits as String: 256, kSecAttrTokenID as String: kSecAttrTokenIDSecureEnclave, kSecAttrAccessControl as String: access as Any, kSecAttrApplicationTag as String: "com.example.key".data(using: .utf8)!, kSecReturnRef as String: true ] let privateKey = SecKeyCreateRandomKey(attributes as CFDictionary, nil) Later, I use this key to sign a message: let signature = SecKeyCreateSignature(privateKey, .ecdsaSignatureMessageX962SHA256, dataToSign as CFData, nil) This prompts for biometric authentication, but shows the default system text. How can I customize or localize the biometric prompt (e.g., title, description, button text) shown during SecKeyCreateSignature? Thanks!

Hello, I have created an app for both iOs and Android where upon speaking two trigger words, the listening app sends a text and then calls to an inputted designated phone contact. The Android version works perfectly. The iOs version also works perfectly but the iOs app emiits a PopUp for each, the text and then the call asking "Do you really want to send the text -or- make the call". Basically, I input the contact info and I spoke the trigger words. So, yes I want to send the text and make the call. So, I have to click the two PopUps then the device sends and calls. Is there a way to suppress the PopUps in any way? The app is designed for emergencies. So, a dely to anser a popup is not at all good. Maybe by telling the device to allow auto calls and texts from my app? Any and all help on this issue will be very welcomed... Thanks :)

Hello, I have a password manager app and would like to help my user's to enable the Safari autofill capability. I've made the password credential extension and that is working great. I just need to help my user's enable the feature. I could point them to Settings-&gt;General-&gt;AutoFill &amp; Password and instruct them to turn the feature on. However, I've noticed that my competitors are able to present an alert directly from their app to turn the feature on (without going to settings at all). I can't find any documentation on how to do this? Thanks for your help! //Ray

I am developing a MacOS Authorisation Plugin, I have username and password entry items and utilising SFAuthorizationPluginView to display that. I am able to do so. Requirement is I have to store ed25519 private key in PEM format in System Keychain as I need to read this entry before login to sign a request to a remote server. I only want my authorisation plugin to access this private key in System Keychain. I am looking up resources on the internet but I could not find specific to macOS Authorisation plugin, many are specific to iOS and some point at using entitlements and app group, but I doubt that applies to macOS authorisation plugin. I'll really appreciate if some pointers are shared how can I store a private credential in System Keychain so that it can be used by only my plugin only, and this is before I have logged into the system.

Hello, I have a password manager app and have noticed a new feature in AutoFill &amp; Passwords called "Set Up Codes In". I see that my competitors have been able to implement this feature but cannot find any documentation on how to do this. How can I make it so my app can support this feature. Any help to pointing me to the documentation or otherwise would be greatly appreciated. Thanks! //Ray

Regarding the issue of login controls remaining on screen for a few seconds when using a subclass of SFAuthorizationPluginView, I wanted to inquire whether any progress has been made on resolving it. To recap, per notes I found in the QAuthPlugins sample code: Due to a bug (FB12074874), the use of an SFAuthorizationPluginView subclass can cause the login controls to remain onscreen for a significant amount of time (roughly 5 seconds) after login is complete, resulting in them being onscreen at the same time as the Finder’s menu bar and the Dock. The exact circumstances under which this happens are not well understood, but one factor seems to be running on a laptop where the main display is mirrored to an external display. Specifically, I would like to know: If there any other information about how the issue is reproduced? For my part I can say that it reproduces with out the use of a mirrored display. So far it reproduces for all of our developers and testers, all of the time. Are there any known workarounds? Is there any expectation that this issue will be addressed? Thank you so much!

Hi, When calling generateAssertion on DCAppAttestService.shared, it gives invalidKey error when there was an update for an offloaded app. The offloading and reinstall always works fine if it is the same version on app store that was offloaded from device, but if there is an update and the app tries to reuse the keyID from previous installation for generateAssertion, attestation service rejects the key with error code 3 (invalid key) for a significant portion of our user. In our internal testing it failed for more than a third of the update attempts. STEPS TO REPRODUCE: install v1 from app store generate key using DCAppAttestService.shared.generateKey Attest this key using DCAppAttestService.shared.attestKey Send the attestation objection to our server and verify with apple servers Generate assertions for network calls to backend using DCAppAttestService.shared.generateAssertion with keyID from step 2 Device offloads the app (manually triggered by user, or automatically by iOS) A new version v2 is published to App Store Use tries to open the app Latest version is download from the App Store App tries to use the keyID from step 2 to generate assertions DCAppAttestService throws invalidKey error (Error Domain=com.apple.devicecheck.error Code=3) Step 7 is critical here, if there is no new version of the app, the reinstalled v1 can reuse the key from step 2 without any issues Is this behaviour expected? Is there any way we can make sure the key is preserved between offloaded app updates? Thanks

Our app uses Face ID to optionally secure access to the app for device owner. This not the new 'Require Face ID' feature of iOS 18 - this is our own custom implementation that has some other related logic for authentication handling. Starting in iOS 18.3.1, starting the app results in multiple Face Id checks being fired - sometimes just a couple but sometimes many more. Curiously, this is happening even when I completely disable any code we have that prompts for Face ID. It appears to come from nowhere. This does not happen on prior iOS 18 releases so, while I might be doing something improper in the code, something specific has changed in iOS 18.3.1 to cause this issue to manifest. I'm looking for advice as to what could be occurring here, how to debug a Face Id check that appears to come from nowhere, and what, if any, workarounds exist.

I am developing a macOS application (targeting macOS 13 and later) that is non-sandboxed and needs to install and trust a root certificate by adding it to the System keychain programmatically. I’m fine with prompting the user for admin privileges or password, if needed. So far, I have attempted to execute the following command programmatically from both: A user-level process A root-level process sudo security add-trusted-cert -d -r trustRoot -k /Library/Keychains/System.keychain /path/to/cert.pem While the certificate does get installed, it does not appear as trusted in the Keychain Access app. One more point: The app is not distributed via MDM. App will be distributed out side the app store. Questions: What is the correct way to programmatically install and trust a root certificate in the System keychain? Does this require additional entitlements, signing, or profile configurations? Is it possible outside of MDM management? Any guidance or working samples would be greatly appreciated.

Hello, I'm seeking some clarification regarding the use of accessibility and input monitoring APIs in sandboxed apps that are distributed through the App Store. I understand that accessibility permissions are generally restricted for App Store apps. However, I've seen several recently released apps request these permissions directly upon first launch. I'm aware that apps submitted prior to 2012 may have legacy access to certain APIs, but the ones I'm referring to appear to be recent - within the past year. While it's possible these apps were approved despite the restrictions, I want to make sure I'm not overlooking something. I also came across a recent discussion on this topic, and one post in particular stood out: Link I’d really appreciate some clarification on what's officially allowed. Specifically: Are accessibility permissions ever allowed? If so, under what circumstances? Is input monitoring permitted for apps on the App Store? (The referenced post says yes, but since it's from 2022, I just want to confirm) The linked post suggests that event generation might be allowed on the App Store, though the author hadn’t explored that privilege in detail and recommended opening a DTS tech support incident. I’ve done that and have a support case open - would it be possible to take a closer look at this? For context, my app (currently distributed outside the App Store) uses CGEventPost and CGEventCreateMouseEvent to modify mouse behavior. Thank you

We’ve noticed an unexpected behavior in our production iOS app where the UIDevice.current.identifierForVendor value occasionally changes, even though: The app is distributed via the App Store (not TestFlight or Xcode builds) We do not switch provisioning profiles or developer accounts No App Clips, App Thinning, or other advanced features are in use There’s no manual reinstall or device reset in the scenarios observed (as per user feedback) Any insights or confirmations would be much appreciated. Thanks!

Using the simplified sign-in with tvOS and a third party password manager, I receive a complete ASPasswordCredential, and I can easily log into my app. When I do the same thing but with Apple's password manager as the source, I receive an ASPasswordCredential that includes the email address, but the password is an empty string. I have tried deleting the credentials from Apple Passwords and regenerating them with a new login to the app's website. I have tried restarting my iPhone. Is this the expected behavior? How should I be getting a password from Apple's Password app with an ASAuthorizationPasswordRequest?