privacyguides/privacyguides.org
1
1
Fork 0
mirror of https://github.com/privacyguides/privacyguides.org.git synced 2026-05-30 15:19:10 +00:00
Code Issues Activity

style: Rename to Wiki section

Browse Source
This commit is contained in:
Jonah Aragon
2026-05-08 23:15:17 -05:00
parent 2a2aabe9dc
commit adb124c488
43 changed files with 390 additions and 327 deletions
Download Patch File Download Diff File Expand all files Collapse all files
content/wiki/os/android-overview.md
+172
View File
@@ -0,0 +1,172 @@
---
title: Android Overview
icon: simple/android
description: Android is an open-source operating system with strong security protections, which makes it our top choice for phones.
robots: nofollow, max-snippet:-1, max-image-preview:large
---
![Android logo](../assets/img/android/android.svg){ align=right }
The **Android Open Source Project** is a secure mobile operating system featuring strong [app sandboxing](https://source.android.com/security/app-sandbox), [Verified Boot](https://source.android.com/security/verifiedboot) (AVB), and a robust [permission](https://developer.android.com/guide/topics/permissions/overview) control system.
[:octicons-home-16:](https://source.android.com){ .card-link title=Homepage }
[:octicons-info-16:](https://source.android.com/docs){ .card-link title=Documentation}
[:octicons-code-16:](https://cs.android.com/android/platform/superproject/main){ .card-link title="Source Code" }
[Our Android Advice :material-arrow-right-drop-circle:](../android/index.md){ .md-button .md-button--primary }
## Security Protections
Key components of the Android security model include [verified boot](#verified-boot), [firmware updates](#firmware-updates), and a robust [permission system](#android-permissions). These important security features form the baseline of the minimum criteria for our [mobile phone](../mobile-phones.md) and [custom Android OS](../android/distributions.md) recommendations.
### Verified Boot
[**Verified Boot**](https://source.android.com/security/verifiedboot) is an important part of the Android security model. It provides protection against [evil maid](https://en.wikipedia.org/wiki/Evil_maid_attack) attacks, malware persistence, and ensures security updates cannot be downgraded with [rollback protection](https://source.android.com/security/verifiedboot/verified-boot#rollback-protection).
Android 10 and above has moved away from full-disk encryption to more flexible [file-based encryption](https://source.android.com/security/encryption/file-based). Your data is encrypted using unique encryption keys, and the operating system files are left unencrypted.
Verified Boot ensures the integrity of the operating system files, thereby preventing an adversary with physical access from tampering or installing malware on the device. In the unlikely case that malware is able to exploit other parts of the system and gain higher privileged access, Verified Boot will prevent and revert changes to the system partition upon rebooting the device.
Unfortunately, OEMs are only obliged to support Verified Boot on their stock Android distribution. Only a few OEMs such as Google support custom AVB key enrollment on their devices. Additionally, some AOSP derivatives such as LineageOS or /e/ OS do not support Verified Boot even on hardware with Verified Boot support for third-party operating systems. We recommend that you check for support **before** purchasing a new device. AOSP derivatives which do not support Verified Boot are **not** recommended.
Many OEMs also have broken implementation of Verified Boot that you have to be aware of beyond their marketing. For example, the Fairphone 3 and 4 are not secure by default, as the [stock bootloader trusts the public AVB signing key](https://forum.fairphone.com/t/bootloader-avb-keys-used-in-roms-for-fairphone-3-4/83448/11). This breaks verified boot on a stock Fairphone device, as the system will boot alternative Android operating systems (such as /e/) [without any warning](https://source.android.com/security/verifiedboot/boot-flow#locked-devices-with-custom-root-of-trust) about custom operating system usage.
### Firmware Updates
**Firmware updates** are critical for maintaining security and without them your device cannot be secure. OEMs have support agreements with their partners to provide the closed-source components for a limited support period. These are detailed in the monthly [Android Security Bulletins](https://source.android.com/security/bulletin).
As the components of the phone, such as the processor and radio technologies rely on closed-source components, the updates must be provided by the respective manufacturers. Therefore, it is important that you purchase a device within an active support cycle. [Qualcomm](https://qualcomm.com/news/releases/2020/12/qualcomm-and-google-announce-collaboration-extend-android-os-support-and) and [Samsung](https://news.samsung.com/us/samsung-galaxy-security-extending-updates-knox) support their devices for 4 years, while cheaper products often have shorter support cycles. With the introduction of the [Pixel 6](https://support.google.com/pixelphone/answer/4457705), Google now makes their own SoC, and they will provide a minimum of 5 years of support. With the introduction of the Pixel 8 series, Google increased that support window to 7 years.
EOL devices which are no longer supported by the SoC manufacturer cannot receive firmware updates from OEM vendors or after market Android distributors. This means that security issues with those devices will remain unfixed.
Fairphone, for example, markets their Fairphone 4 device as receiving 6 years of support. However, the SoC (Qualcomm Snapdragon 750G on the Fairphone 4) has a considerably shorter EOL date. This means that firmware security updates from Qualcomm for the Fairphone 4 will end in September 2023, regardless of whether Fairphone continues to release software security updates.
### Android Permissions
[**Permissions on Android**](https://developer.android.com/guide/topics/permissions/overview) grant you control over what apps are allowed to access. Google regularly makes [improvements](https://developer.android.com/about/versions/11/privacy/permissions) on the permission system in each successive version. All apps you install are strictly [sandboxed](https://source.android.com/security/app-sandbox), therefore, there is no need to install any antivirus apps.
A smartphone with the latest version of Android will always be more secure than an old smartphone with an antivirus that you have paid for. It's better not to pay for antivirus software and to save money to buy a new smartphone such as a [Google Pixel](../mobile-phones.md#google-pixel).
Android 10:
- [Scoped Storage](https://developer.android.com/about/versions/10/privacy/changes#scoped-storage) gives you more control over your files and can limit what can [access external storage](https://developer.android.com/training/data-storage#permissions). Apps can have a specific directory in external storage as well as the ability to store specific types of media there.
- Tighter access on [device location](https://developer.android.com/about/versions/10/privacy/changes#app-access-device-location) by introducing the `ACCESS_BACKGROUND_LOCATION` permission. This prevents apps from accessing the location when running in the background without express permission from the user.
Android 11:
- [One-time permissions](https://developer.android.com/about/versions/11/privacy/permissions#one-time) which allows you to grant a permission to an app just once.
- [Auto-reset permissions](https://developer.android.com/about/versions/11/privacy/permissions#auto-reset), which resets [runtime permissions](https://developer.android.com/guide/topics/permissions/overview#runtime) that were granted when the app was opened.
- Granular permissions for accessing [phone number](https://developer.android.com/about/versions/11/privacy/permissions#phone-numbers) related features.
Android 12:
- A permission to grant only the [approximate location](https://developer.android.com/about/versions/12/behavior-changes-12#approximate-location).
- Auto-reset of [hibernated apps](https://developer.android.com/about/versions/12/behavior-changes-12#app-hibernation).
- [Data access auditing](https://developer.android.com/about/versions/12/behavior-changes-12#data-access-auditing) which makes it easier to determine what part of an app is performing a specific type of data access.
Android 13:
- A permission for [nearby Wi-Fi access](https://developer.android.com/about/versions/13/behavior-changes-13#nearby-wifi-devices-permission). The MAC addresses of nearby Wi-Fi access points were a popular way for apps to track a user's location.
- More [granular media permissions](https://developer.android.com/about/versions/13/behavior-changes-13#granular-media-permissions), meaning you can grant access to images, videos or audio files only.
- Background use of sensors now requires the [`BODY_SENSORS`](https://developer.android.com/about/versions/13/behavior-changes-13#body-sensors-background-permission) permission.
An app may request a permission for a specific feature it has. For example, any app that can scan QR codes will require the camera permission. Some apps can request more permissions than they need.
[Exodus](https://exodus-privacy.eu.org) can be useful when comparing apps that have similar purposes. If an app requires a lot of permissions and has a lot of advertising and analytics this is probably a bad sign. We recommend looking at the individual trackers and reading their descriptions rather than simply **counting the total** and assuming all items listed are equal.
<div class="admonition warning" markdown>
<p class="admonition-title">Warning</p>
If an app is mostly a web-based service, the tracking may occur on the server side. [Facebook](https://reports.exodus-privacy.eu.org/en/reports/com.facebook.katana/latest) shows "no trackers" but certainly does track users' interests and behavior across the site. Apps may evade detection by not using standard code libraries produced by the advertising industry, though this is unlikely.
</div>
<div class="admonition note" markdown>
<p class="admonition-title">Note</p>
Privacy-friendly apps such as [Bitwarden](https://reports.exodus-privacy.eu.org/en/reports/com.x8bit.bitwarden/latest) may show some trackers such as [Google Firebase Analytics](https://reports.exodus-privacy.eu.org/en/trackers/49). This library includes [Firebase Cloud Messaging](https://en.wikipedia.org/wiki/Firebase_Cloud_Messaging) which can provide [push notifications](https://en.wikipedia.org/wiki/Push_technology) in apps. This [is the case](https://fosstodon.org/@bitwarden/109636825700482007) with Bitwarden. That doesn't mean that Bitwarden is using all the analytics features that are provided by Google Firebase Analytics.
</div>
## Privacy Features
### User Profiles
Multiple **user profiles** can be found in :gear: **Settings** → **System** → **Users** and are the simplest way to isolate in Android.
With user profiles, you can impose restrictions on a specific profile, such as: making calls, using SMS, or installing apps. Each profile is encrypted using its own encryption key and cannot access the data of any other profiles. Even the device owner cannot view the data of other profiles without knowing their password. Multiple user profiles are a more secure method of isolation.
### Work Profile
[**Work Profiles**](https://support.google.com/work/android/answer/6191949) are another way to isolate individual apps and may be more convenient than separate user profiles.
A **device controller** app such as [Shelter](../android/general-apps.md#shelter) is required to create a Work Profile without an enterprise MDM, unless you're using a custom Android OS which includes one.
The work profile is dependent on a device controller to function. Features such as *File Shuttle* and *contact search blocking* or any kind of isolation features must be implemented by the controller. You must also fully trust the device controller app, as it has full access to your data inside the work profile.
This method is generally less secure than a secondary user profile; however, it does allow you the convenience of running apps in both the owner profile and work profile simultaneously.
### Private Space
**Private Space** is a feature introduced in Android 15 that adds another way of isolating individual apps. You can set up a private space in the owner profile by navigating to :gear: **Settings** → **Security & privacy** → **Private space**. Once set up, your private space resides at the bottom of the app drawer.
Like user profiles, a private space is encrypted using its own encryption key, and you have the option to set up a different unlock method. Like work profiles, you can use apps from both the owner profile and private space simultaneously. Apps launched from a private space are distinguished by an icon depicting a key within a shield.
Unlike work profiles, Private Space is a feature native to Android that does not require a third-party app to manage it. For this reason, we generally recommend using a private space over a work profile, though you can use a work profile alongside a private space.
### VPN kill switch
Android 7 and above supports a VPN kill switch, and it is available without the need to install third-party apps. This feature can prevent leaks if the VPN is disconnected. It can be found in :gear: **Settings** → **Network & internet** → **VPN** → :gear: → **Block connections without VPN**.
### Global Toggles
Modern Android devices have global toggles for disabling Bluetooth and location services. Android 12 introduced toggles for the camera and microphone. When not in use, we recommend disabling these features. Apps cannot use disabled features (even if granted individual permissions) until re-enabled.
## Google Services
If you are using a device with Google services—whether with the stock operating system or an operating system that safely sandboxes Google Play Services like GrapheneOS—there are a number of additional changes you can make to improve your privacy. We still recommend avoiding Google services entirely, or limiting Google Play Services to a specific user/work profile by combining a device controller like *Shelter* with GrapheneOS's Sandboxed Google Play.
### Advanced Protection Program
If you have a Google account we suggest enrolling in the [Advanced Protection Program](https://landing.google.com/advancedprotection). It is available at no cost to anyone with two or more hardware security keys with [FIDO](../basics/multi-factor-authentication.md#fido-fast-identity-online) support. Alternatively, you can use [passkeys](https://fidoalliance.org/passkeys).
The Advanced Protection Program provides enhanced threat monitoring and enables:
- Stricter two-factor authentication; e.g. that [FIDO](../basics/multi-factor-authentication.md#fido-fast-identity-online) **must** be used and disallows the use of [SMS OTPs](../basics/multi-factor-authentication.md#sms-or-email-mfa), [TOTP](../basics/multi-factor-authentication.md#time-based-one-time-password-totp) and [OAuth](../basics/account-creation.md#sign-in-with-oauth)
- Only Google and verified third-party apps can access account data
- Scanning of incoming emails on Gmail accounts for [phishing](https://en.wikipedia.org/wiki/Phishing#Email_phishing) attempts
- Stricter [safe browser scanning](https://google.com/chrome/privacy/whitepaper.html#malware) with Google Chrome
- Stricter recovery process for accounts with lost credentials
If you use non-sandboxed Google Play Services (common on stock operating systems), the Advanced Protection Program also comes with [additional benefits](https://support.google.com/accounts/answer/9764949) such as:
- Not allowing app installation outside the Google Play Store, the OS vendor's app store, or via [`adb`](https://en.wikipedia.org/wiki/Android_Debug_Bridge)
- Mandatory automatic device scanning with [Play Protect](https://support.google.com/googleplay/answer/2812853?#zippy=%2Chow-malware-protection-works%2Chow-privacy-alerts-work)
- Warning you about unverified applications
- Enabling ARM's hardware-based [Memory Tagging Extension (MTE)](https://developer.arm.com/documentation/108035/0100/Introduction-to-the-Memory-Tagging-Extension) for supported apps, which lowers the likelihood of device exploits happening through memory corruption bugs
### Google Play System Updates
In the past, Android security updates had to be shipped by the operating system vendor. Android has become more modular beginning with Android 10, and Google can push security updates for **some** system components via the privileged Play Services.
If you have an EOL device shipped with Android 10 or above and are unable to run any of our recommended operating systems on your device, you are likely going to be better off sticking with your OEM Android installation (as opposed to an operating system not listed here such as LineageOS or /e/ OS). This will allow you to receive **some** security fixes from Google, while not violating the Android security model by using an insecure Android derivative and increasing your attack surface. We would still recommend upgrading to a supported device as soon as possible.
### Advertising ID
All devices with Google Play Services installed automatically generate an [advertising ID](https://support.google.com/googleplay/android-developer/answer/6048248) used for targeted advertising. Disable this feature to limit the data collected about you.
On Android distributions with [sandboxed Google Play](https://grapheneos.org/usage#sandboxed-google-play), go to :gear: **Settings** → **Apps** → **Sandboxed Google Play** → **Google Settings** → **All services** → **Ads**.
- [x] Select **Delete advertising ID**
On Android distributions with privileged Google Play Services (which includes the stock installation on most devices), the setting may be in one of several locations. Check
- :gear: **Settings** → **Google** → **Ads**
- :gear: **Settings** → **Privacy** → **Ads**
You will either be given the option to delete your advertising ID or to *Opt out of interest-based ads* (this varies between OEM distributions of Android). If presented with the option to delete the advertising ID, that is preferred. If not, then make sure to opt out and reset your advertising ID.
### SafetyNet and Play Integrity API
[SafetyNet](https://developer.android.com/training/safetynet/attestation) and the [Play Integrity APIs](https://developer.android.com/google/play/integrity) are generally used for [banking apps](https://grapheneos.org/usage#banking-apps). Many banking apps will work fine in GrapheneOS with sandboxed Play services, however some non-financial apps have their own crude anti-tampering mechanisms which might fail. GrapheneOS passes the `basicIntegrity` check, but not the certification check `ctsProfileMatch`. Devices with Android 8 or later have hardware attestation support which cannot be bypassed without leaked keys or serious vulnerabilities.
As for Google Wallet, we don't recommend this due to their [privacy policy](https://payments.google.com/payments/apis-secure/get_legal_document?ldo=0&ldt=privacynotice&ldl=en), which states you must opt out if you don't want your credit rating and personal information shared with affiliate marketing services.
content/wiki/os/index.md
+22
View File
@@ -0,0 +1,22 @@
---
title: Operating Systems
description: An overview of our operating system-related recommendations for all major computing hardware.
---
We publish configuration guides for the major operating systems, because you can generally improve the amount of data that is collected about you on any option, especially if you use privacy tools like our [recommended web browsers](../desktop-browsers.md) in place of native tools where appropriate. However, some operating systems will be more privacy-respecting inherently, and it will be much harder to achieve an equivalent level of privacy on other choices.
- [Recommended Linux Distros :material-arrow-right-drop-circle:](../desktop.md)
- [Recommended Android Distros :material-arrow-right-drop-circle:](../android/distributions.md)
The articles marked with a :material-star: are our more mature articles.
## Mobile Operating Systems
- [Android Overview](android-overview.md) :material-star:
- [iOS Overview](ios-overview.md)
## Desktop Operating Systems
- [Linux Overview](linux-overview.md) :material-star:
- [macOS Overview](macos-overview.md)
- [Qubes Overview](qubes-overview.md) :material-star:
- [Windows Overview](windows/index.md)
content/wiki/os/ios-overview.md
+311
View File
@@ -0,0 +1,311 @@
---
title: iOS Overview
icon: simple/apple
description: iOS is a mobile operating system developed by Apple for the iPhone.
---
**iOS** and **iPadOS** are proprietary mobile operating systems developed by Apple for their iPhone and iPad products, respectively. If you have an Apple mobile device, you can increase your privacy by disabling some built-in telemetry features, and hardening some privacy and security settings which are built in to the system.
## Privacy Notes
iOS devices are frequently praised by security experts for their robust data protection and adherence to modern best practices. However, the restrictiveness of Apple's ecosystem—particularly with their mobile devices—does still hamper privacy in a number of ways.
We generally consider iOS to provide better than average privacy and security protections for most people, compared to stock Android devices from any manufacturer. However, you can achieve even higher standards of privacy with a [custom Android operating system](../android/distributions.md) like GrapheneOS, if you want or need to be completely independent of Apple or Google's cloud services.
### Activation Lock
All iOS devices must be checked against Apple's Activation Lock servers when they are initially set up or reset, meaning an internet connection is **required** to use an iOS device.
### Mandatory App Store
The only source for apps on iOS is Apple's App Store, which requires an Apple Account to access. This means that Apple has a record of every app you install on your device, and can likely tie that information to your actual identity if you provide the App Store with a payment method.
### Invasive Telemetry
Apple has historically had problems with properly disassociating their telemetry from Apple Accounts on iOS. In [2019](https://theguardian.com/technology/2019/jul/26/apple-contractors-regularly-hear-confidential-details-on-siri-recordings), Apple was found to transmit Siri recordings—some containing highly confidential information—to their servers for manual review by third-party contractors. Though Apple temporarily stopped that program after that practice was [widely reported on](https://theverge.com/2019/8/23/20830120/apple-contractors-siri-recordings-listening-1000-a-day-globetech-microsoft-cortana), the company rolled out a switch to [**opt out** of uploading conversations with Siri](https://theguardian.com/technology/2019/oct/30/apple-lets-users-opt-out-of-having-siri-conversations-recorded) a few months later in the succeeding iOS update. Moreover, in 2021, [Apple reworked Siri](https://theguardian.com/technology/2021/jun/07/apple-overhauls-siri-to-address-privacy-concerns-and-improve-performance) so that it processes voice recordings locally rather than sending it to their servers.
More recently, Apple has been found to transmit analytics [even when analytics sharing is disabled](https://gizmodo.com/apple-iphone-analytics-tracking-even-when-off-app-store-1849757558) on iOS, and this data [appears](https://twitter.com/mysk_co/status/1594515229915979776) to be easily linked to unique iCloud account identifiers despite supposedly being decoupled from Apple Accounts.
### Traffic Outside Active VPN Connections
Apple's [privacy policy regarding VPNs](https://apple.com/legal/privacy/data/en/vpns) states:
> Even when a VPN is active, some traffic that is necessary for essential system services will take place outside the VPN so that your device can function properly.
## Recommended Configuration
**Note:** This guide assumes that you're running the latest version of iOS.
### iCloud
The majority of privacy and security concerns with Apple products are related to their cloud services, not their hardware or software. When you use Apple services like iCloud, most of your information is stored on their servers and secured with keys which Apple has access to by default. You can check [Apple's documentation](https://support.apple.com/HT202303) for information on which services are end-to-end encrypted. Anything listed as "in transit" or "on server" means it's possible for Apple to access that data without your permission. This level of access has occasionally been abused by law enforcement to get around the fact that your data is otherwise securely encrypted on your device, and of course Apple is vulnerable to data breaches like any other company.
Therefore, if you do use iCloud you should [enable **Advanced Data Protection**](https://support.apple.com/HT212520). This encrypts nearly all of your iCloud data with keys stored on your devices (end-to-end encryption), rather than Apple's servers, so that your iCloud data is secured in the event of a data breach, and otherwise hidden from Apple.
The encryption used by Advanced Data Protection, while strong, [is not *quite* as robust](https://discuss.privacyguides.net/t/apple-advances-user-security-with-powerful-new-data-protections/10778/4) as the encryption offered by other [cloud services](../cloud.md), particularly when it comes to iCloud Drive. While we strongly encourage using Advanced Data Protection if you use iCloud, we would also suggest considering finding an alternative to iCloud from a more [privacy-focused service provider](../tools.md), although it is unlikely most people would be impacted by these encryption quirks.
You can also protect your data by limiting what you sync to iCloud in the first place. At the top of the **Settings** app, you'll see your name and profile picture if you are signed in to iCloud. Select that, then **iCloud**, and turn off the switches for any services you don't want to sync to iCloud. You may see third-party apps listed under **Show All** if they sync to iCloud, which you can disable here.
#### iCloud+
A paid **iCloud+** subscription (with any iCloud storage plan) comes with some privacy-protecting functionality. While these may provide adequate service for current iCloud customers, we wouldn't recommend purchasing an iCloud+ plan over a [VPN](../vpn.md) and [standalone email aliasing service](../email-aliasing.md) just for these features alone.
[**Private Relay**](https://apple.com/legal/privacy/data/en/icloud-relay) is a proxy service which relays all of your Safari traffic, your DNS queries, and unencrypted traffic on your device through two servers: one owned by Apple and one owned by a third-party provider (including Akamai, Cloudflare, and Fastly). In theory this should prevent any single provider in the chain—including Apple—from having full visibility into which websites you visit while connected. Unlike a VPN, Private Relay does not protect traffic that's already encrypted.
**Hide My Email** is Apple's email aliasing service. You can create an email aliases for free when you *Sign In With Apple* on a website or app, or generate unlimited aliases on demand with a paid iCloud+ plan. Hide My Email has the advantage of using the `@icloud.com` domain for its aliases, which may be less likely to be blocked compared to other email aliasing services, but does not offer functionality offered by standalone services such as automatic PGP encryption or multiple mailbox support.
#### Media & Purchases
At the top of the **Settings** app, you'll see your name and profile picture if you are signed in to an Apple Account. Select that, then select **Media & Purchases** → **View Account**.
- [ ] Turn off **Personalized Recommendations**
#### Find My
**Find My** is a service that lets you track your Apple devices and share your location with your friends and family. It also allows you to wipe your device remotely in case it is stolen, preventing a thief from accessing your data. Your Find My [location data is E2EE](https://apple.com/legal/privacy/data/en/find-my) when:
- Your location is shared with a family member or friend, and you both use iOS 17 or greater.
- Your device is offline and is located by the Find My Network.
Your location data is not E2EE when your device is online and you use Find My iPhone remotely to locate your device. You will have to make the decision whether these trade-offs are worth the anti-theft benefits of Activation Lock.
At the top of the **Settings** app, you'll see your name and profile picture if you are signed in to an Apple Account. Select that, then select **Find My**. Here you can choose whether to enable or disable Find My location features.
### Settings
Many other privacy-related settings can be found in the **Settings** app.
#### Airplane Mode
Enabling **Airplane Mode** stops your phone from contacting cell towers. You will still be able to connect to Wi-Fi and Bluetooth, so whenever you are connected to Wi-Fi you can turn this setting on.
#### Wi-Fi
You can enable [hardware address randomization](https://support.apple.com/en-us/102509#triswitch) to protect you from tracking across Wi-Fi networks, and on the same network over time. On the network you are currently connected to, tap the :material-information: button:
- [x] Set **Private Wi-Fi Address** to **Fixed** or **Rotating**
You also have the option to **Limit IP Address Tracking**. This is similar to iCloud Private Relay but only affects connections to "known trackers." Because it only affects connections to potentially malicious servers, this setting is probably fine to leave enabled, but if you don't want *any* traffic to be routed through Apple's servers, you should turn it off.
#### Bluetooth
**Bluetooth** should be disabled when you aren't using it as it increases your attack surface. Disabling Bluetooth (or Wi-Fi) via the Control Center only disables it temporarily: you must switch it off in Settings for disabling it to remain effective.
- [ ] Turn off **Bluetooth**
Note that Bluetooth is automatically turned on after every system update.
#### General
Your iPhone's device name will by default contain your first name, and this will be visible to anyone on networks you connect to. You should change this to something more generic, like "iPhone." Select **About** → **Name** and enter the device name you prefer.
It is important to install software updates frequently to get the latest security fixes. You can enable automatic updates to keep your phone up-to-date without needing to constantly check for updates. Select **Software Update** → **Automatic Updates**:
- [x] Turn on **Automatically Install**
**AirDrop** is commonly used to easily share files, but it represents a significant privacy risk. The AirDrop protocol constantly broadcasts your personal information to your surroundings, with [very weak](https://usenix.org/system/files/sec21-heinrich.pdf) security protections. Your identity can easily be discovered by attackers even with limited resources, and the Chinese government has [openly acknowledged](https://arstechnica.com/security/2024/01/hackers-can-id-unique-apple-airdrop-users-chinese-authorities-claim-to-do-just-that) using such techniques to identify AirDrop users in public since 2022.
- [x] Select **AirDrop** → **Receiving Off**
**AirPlay** lets you seamlessly stream content from your iPhone to a TV; however, you might not always want this. Select **AirPlay & Continuity** → **Automatically AirPlay**:
- [x] Select **Never** or **Ask**
**Background App Refresh** allows your apps to refresh their content while you're not using them. This may cause them to make unwanted connections. Turning this off can also save battery life, but may affect an app's ability to receive updated information, particularly weather and messaging apps.
Select **Background App Refresh** and switch off any apps you don't want to continue refreshing in the background. If you don't want any apps to refresh in the background, you can select **Background App Refresh** again and turn it **Off**.
#### Apple Intelligence & Siri
This is available if your device supports **[Apple Intelligence](https://support.apple.com/guide/iphone/apple-intelligence-and-privacy-iphe3f499e0e/ios)**. Apple Intelligence uses a combination of on-device processing and their **[Private Cloud Compute](https://security.apple.com/blog/private-cloud-compute)** for things that take more processing power than your device can provide.
To see a report of all the requests made to Apple's servers, you can navigate to **Privacy & Security** → **Apple Intelligence Report** and press **Export Activity** to see activity from the either the last 15 minutes or 7 days, depending on what you set it for. Similar to the **App Privacy Report** which shows you the recent permissions accessed by the apps on your phone, the Apple Intelligence Report likewise shows what is being sent to Apple's servers while using Apple Intelligence.
Apple Intelligence can integrate with [ChatGPT](https://support.apple.com/guide/iphone/use-chatgpt-with-apple-intelligence-iph00fd3c8c2/ios). If you want ChatGPT integration, you can navigate to **ChatGPT** and press **Set Up**. If you want to disable it, go to the same place:
- [ ] Turn off **Use ChatGPT**
You can also have it ask for confirmation every time if you leave ChatGPT integration on:
- [x] Turn on **Confirm Requests**
If you don't want anyone to be able to control your phone with Siri when it is locked, you can turn that off here.
- [ ] Turn off **Allow Siri When Locked**
#### Face ID/Touch ID & Passcode
Setting a strong password on your phone is the most important step you can take for physical device security. You'll have to make trade-offs here between security and convenience: A longer password will be annoying to type in every time, but a shorter password or PIN will be easier to guess. Setting up Face ID or Touch ID along with a strong password can be a good compromise between usability and security.
Select **Turn Passcode On** or **Change Passcode** → **Passcode Options** → **Custom Alphanumeric Code**. Make sure that you create a [secure password](../basics/passwords-overview.md).
If you wish to use Face ID or Touch ID, you can go ahead and set it up now. Your phone will use the password you set up earlier as a fallback in case your biometric verification fails. Biometric unlock methods are primarily a convenience, although they do stop surveillance cameras or people over your shoulder from watching you input your passcode.
If you use biometrics, you should know how to turn them off quickly in an emergency. Holding down the [side button](https://support.apple.com/en-us/105103) and *either* volume button until you see the Slide to Power Off slider will disable biometrics, requiring your passcode to unlock. Your passcode will be required after your device restarts.
You can similarly disable biometrics by pressing the side button five times, or for devices with Touch ID, you can hold down the side button and nothing else. Make sure you try this in advance, so you know which method works for your device.
**Stolen Device Protection** adds additional security intended to protect your personal data if your device is stolen while unlocked. If you enable both biometric authentication and the [Find My](#find-my) iPhone feature, we recommend enabling this protection:
- [x] Turn on **Stolen Device Protection**
After enabling Stolen Device Protection, [certain actions](https://support.apple.com/HT212510) will require biometric authentication without a password fallback (in the event that a shoulder surfer has obtained your PIN), such as using password autofill, accessing payment information, and disabling Lost Mode. It also adds a security delay to certain actions performed away from your home or another "familiar location," such as requiring a 1-hour timer to reset your Apple Account password or sign out of your Apple Account. This delay is intended to give you time to enable Lost Mode and secure your account before a thief can reset your device.
**Allow Access When Locked** presents options for what you can allow when your phone is locked. Pick and choose which feature you want to disable to prevent unauthorized access if someone gets their hands on your phone. The more of these options you disable, the less someone without your password can do, but the less convenient it will be for you.
iPhones are already resistant to brute-force attacks by making you wait long periods of time after multiple failed attempts; however, there have historically been exploits to get around this. To be extra safe, you can set your phone to wipe itself after 10 failed passcode attempts.
<div class="admonition warning" markdown>
<p class="admonition-title">Warning</p>
With this setting enabled, someone could intentionally wipe your phone by entering the wrong password many times. Make sure you have proper backups and only enable this setting if you feel comfortable with it.
</div>
- [x] Turn on **Erase Data**
#### Privacy & Security
**Location Services** allows you to use features like Find My and Maps. If you don't need these features, you can disable Location Services. Alternatively, you can review and pick which apps can use your location here. Select **Location Services**:
- [ ] Turn off **Location Services**
A purple arrow will appear next to an app in these settings that has used your location recently, while a gray arrow indicates that your location has been accessed within the last 24 hours. If you decide to leave Location Services on, Apple will use it for System Services by default. You can review and pick which services can use your location here. However, if you don't want to submit location analytics to Apple, which they use to improve Apple Maps, you can disable this here as well. Select **System Services**:
- [ ] Turn off **iPhone Analytics**
- [ ] Turn off **Routing & Traffic**
- [ ] Turn off **Improve Maps**
You can decide to allow apps to request to **track** you here. Disabling this disallows all apps from tracking you with your phone's advertising ID. Select **Tracking**:
- [ ] Turn off **Allow Apps to Request to Track**
This is disabled by default and cannot be changed for users under 18.
You should turn off **Research Sensor & Usage Data** if you don't wish to participate in studies. Select **Research Sensor & Usage Data**:
- [ ] Turn off **Sensor & Usage Data Collection**
**[Safety Check](https://support.apple.com/guide/personal-safety/safety-check-iphone-ios-16-ips2aad835e1/1.0/web/1.0)** allows you to quickly view and revoke certain people and apps that might have permission to access your data. Here, you can perform an **Emergency Reset**, immediately resetting permissions for all people and apps which might have access to device resources. You can also **Manage Sharing & Access**, which allows you to review and customize who and what has access to your device and account resources. If you're in an abusive situation, read Apple's [Personal Safety User Guide](https://support.apple.com/guide/personal-safety/welcome/web) for guidance on what you should do.
You should disable analytics if you don't wish to send usage data to Apple. Select **Analytics & Improvements** and unselect the type(s) of analytics that you don't want to send to Apple.
Disable **Personalized Ads** if you don't want targeted ads. Select **Apple Advertising**:
- [ ] Turn off **Personalized Ads**
**App Privacy Report** is a built-in tool that allows you to see which permissions your apps are using. Select **App Privacy Report**:
- [x] Select **Turn On App Privacy Report**
Set wired accessories to ask for permission when you connect them. Select **Wired Accessories**:
- [x] Select **Always Ask** or **Ask for New Accessories**
**[Lockdown Mode](https://blog.privacyguides.org/2022/10/27/macos-ventura-privacy-security-updates/#lockdown-mode)** is a security setting you can enable to make your phone more resistant to attacks. Be aware that certain apps and features [won't work](https://support.apple.com/HT212650) as they do normally.
- [x] Select **Turn On Lockdown Mode**
## Additional Advice
### E2EE Calls
Normal phone calls made with the Phone app through your carrier are not E2EE. Both FaceTime Video and FaceTime Audio calls are E2EE. Alternatively, you can use [another app](../real-time-communication.md) like Signal for E2EE calls.
### Encrypted iMessage
The [color of the message bubble](https://support.apple.com/en-us/104972) in the Messages app indicates whether your messages are E2EE or not. A blue bubble indicates that you're using iMessage with E2EE, while a green bubble indicates the other party is using either the outdated SMS and MMS protocols or RCS. RCS on iOS is **not** E2EE. Currently, the only way to have E2EE in Messages is for both parties to be using iMessage on Apple devices.
If either you or your messaging partner have iCloud Backup enabled without Advanced Data Protection, the encryption key will be stored on Apple's servers, meaning they can access your messages.
By default, you trust Apple's identity servers that you're messaging the right person. To defend yourself from a potentially malicious server, you can enable **[Contact Key Verification](https://support.apple.com/en-us/118246)**. At the top of the **Settings** app where your name is, select it, then go to **Contact Key Verification**.
- [x] Turn on **Verification in iMessage**
Both you and your contacts need to enable Contact Key Verification and follow Apple's [instructions](https://support.apple.com/en-us/118246#verify) for the security assurances mentioned above to take effect.
### Photo Permissions
When an app prompts you for access to your device's photo library, iOS provides you with options to limit what an app can access.
Rather than allow an app to access all the photos on your device, you can allow it to only access whichever photos you choose by tapping the "Select Photos..." option in the permission dialog. You can change photo access permissions at any time by navigating to **Settings** → **Privacy & Security** → **Photos**.
![Photo Permissions](../assets/img/ios/photo-permissions-light.png#only-light) ![Photo Permissions](../assets/img/ios/photo-permissions-dark.png#only-dark)
**Add Photos Only** is a permission that only gives an app the ability to download photos to the photo library. Not all apps which request photo library access provide this option.
![Private Access](../assets/img/ios/private-access-light.png#only-light) ![Private Access](../assets/img/ios/private-access-dark.png#only-dark)
Some apps also support **Private Access**, which functions similarly to the **Limited Access** permission. However, photos shared to apps using Private Access include their location by default. We recommend unchecking this setting if you do not [remove photo metadata](../data-redaction.md) beforehand.
### Contact Permissions
Similarly, rather than allow an app to access all the contacts saved on your device, you can allow it to only access whichever contacts you choose. You can change contact access permissions at any time by navigating to **Settings** → **Privacy & Security** → **Contacts**.
![Contact Permissions](../assets/img/ios/contact-permissions-light.png#only-light) ![Contact Permissions](../assets/img/ios/contact-permissions-dark.png#only-dark)
### Require Biometrics and Hide Apps
iOS offers the ability to lock most apps behind Touch ID/Face ID or your passcode, which can be useful for protecting sensitive content in apps which do not provide the option themselves. You can lock an app by long-pressing on it and selecting **Require Face ID/Touch ID**. Any app locked in this way requires biometric authentication whenever opening it or accessing its contents in other apps. Also, notification previews for locked apps will not be shown.
In addition to locking apps behind biometrics, you can also hide apps so that they don't appear on the Home Screen, App Library, the app list in **Settings**, etc. While hiding apps may be useful in situations where you have to hand your unlocked phone to someone else, the concealment provided by the feature is not absolute, as a hidden app is still visible in some places such as the battery usage list. Moreover, one notable trade off of hiding an app is that you will not receive any of its notifications.
You can hide an app by long-pressing on it and selecting **Require Face ID/Touch ID** → **Hide and Require Face ID/Touch ID**. Note that pre-installed Apple apps, as well as the default web browser and email app, cannot be hidden. Hidden apps reside in a **Hidden** folder at the bottom of the App Library, which can be unlocked using biometrics. This folder appears in the App Library whether you hid any apps or not, which provides you a degree of plausible deniability.
### Guided Access
Sometimes you might want to hand your phone to someone to make a call or do a specific task, but you don't want them to have full access to your phone. In these cases, you can quickly enable **[Guided Access](https://support.apple.com/guide/iphone/lock-iphone-to-one-app-iph7fad0d10/ios)** to lock the phone to one specific app until you authenticate.
<div class="admonition warning" markdown>
<p class="admonition-title">Warning</p>
Guided Access isn't foolproof, as it's possible you could leak data unintentionally or the feature could be bypassed. You should only use Guided Access for situations where you casually hand your phone to someone to use. You should not use it as a tool to protect against advanced adversaries.
</div>
### Redacting Elements in Images
If you need to hide information in a photo, you can use Apple's built-in editing tools to do so.
You can use the [Clean Up](https://support.apple.com/en-us/121429) feature on supported devices to pixelate faces or remove objects from images.
- Open the **Photos** app and tap the photo you have selected for redaction
- Tap the :material-tune:
- Tap the button labeled **Clean Up**
- Draw a circle around whatever you want to redact. Faces will be pixelated, and it will attempt to delete anything else.
Our warning [against blurring text](../data-redaction.md) also applies here, so we recommend to instead add a black shape with 100% opacity over it. In addition to redacting text, you can also black out any face or object using the **Photos** app.
<div class="annotate" markdown>
- Tap the image you have selected for redaction
- Tap the :material-tune: → :material-dots-horizontal: (1) → Markup → :material-plus:
- Select **Add Shape** and choose the square or circle
- On the toolbar, tap the circle and choose black as the color for filling in the shape. You can also move the shape and increase its size as you see fit.
</div>
1. This may not appear on certain iPhone models.
**Don't** use the highlighter to obfuscate information, as its opacity is not quite 100%.
### Avoid Jailbreaking
Jailbreaking an iPhone undermines its security and makes you vulnerable. Running untrusted, third-party software could cause your device to be infected with malware.
### iOS Betas
Apple always makes beta versions of iOS available early for those that wish to help find and report bugs. We don't recommend installing beta software on your phone. Beta releases are potentially unstable and could have undiscovered security vulnerabilities.
## Security Highlights
### Before First Unlock
If your threat model includes [:material-target-account: Targeted Attacks](../basics/common-threats.md#attacks-against-specific-individuals){ .pg-red } that involve forensic tools, and you want to minimize the chance of exploits being used to access your phone, you should restart your device frequently. The state *after* a reboot but *before* unlocking your device is referred to as "Before First Unlock" (BFU), and when your device is in that state it makes it [significantly more difficult](https://belkasoft.com/checkm8_glossary) for forensic tools to exploit vulnerabilities to access your data. This BFU state allows you to receive notifications for calls, texts, and alarms, but most of the data on your device is still encrypted and inaccessible. This can be impractical, so consider whether these trade-offs make sense for your situation.
iPhones [automatically reboot](https://support.apple.com/guide/security/protecting-user-data-in-the-face-of-attack-secf5549a4f5/1/web/1#:~:text=On%20an%20iPhone%20or%20iPad%20with%20iOS%2018%20and%20iPadOS%2018%20or%20later%2C%20a%20new%20security%20protection%20will%20restart%20devices%20if%20they%20remain%20locked%20for%20a%20prolonged%20period%20of%20time.) if they're not unlocked after a period of time.
### MTE
The iPhone 17 line and later offer a security enhancement called [Memory Tagging Extension](https://developer.arm.com/documentation/108035/0100/Introduction-to-the-Memory-Tagging-Extension) (MTE), which makes it significantly harder for an attacker to exploit memory corruption vulnerabilities. This always-on protection depends on hardware support, so it's not available for older devices.
For more details on Apple's implementation of MTE, read the [blog post](https://security.apple.com/blog/memory-integrity-enforcement) published by Apple Security Research. We also cover Apple's implementation of MTE and how it compares to Android's implementation in the Google Pixel 8 series and later in our [own article](https://www.privacyguides.org/posts/2025/09/20/memory-integrity-enforcement-changes-the-game-on-ios).
content/wiki/os/linux-overview.md
+169
View File
@@ -0,0 +1,169 @@
---
title: Linux Overview
icon: simple/linux
description: Linux is an open-source, privacy-focused desktop operating system alternative, but not all distribitions are created equal.
---
**Linux** is an open-source, privacy-focused desktop operating system alternative. In the face of pervasive telemetry and other privacy-encroaching technologies in mainstream operating systems, desktop Linux has remained the clear choice for people looking for total control over their computers from the ground up.
Our website generally uses the term “Linux” to describe **desktop** Linux distributions. Other operating systems which also use the Linux kernel such as ChromeOS, Android, and Qubes OS are not discussed on this page.
[Our Linux Recommendations :material-arrow-right-drop-circle:](../desktop.md){ .md-button }
## Security Notes
There are some notable security concerns with Linux which you should be aware of. Despite these drawbacks, desktop Linux distributions are still great for most people who want to:
- Avoid telemetry that often comes with proprietary operating systems
- Maintain [software freedom](https://gnu.org/philosophy/free-sw.en.html#four-freedoms)
- Use privacy-focused systems such as [Whonix](../desktop.md#whonix) or [Tails](../desktop.md#tails)
### Open-Source Security
It is a [common misconception](../basics/common-misconceptions.md#open-source-software-is-always-secure-or-proprietary-software-is-more-secure) that Linux and other open-source software are inherently secure simply because the source code is available. There is an expectation that community verification occurs regularly, but this isn’t always [the case](https://seirdy.one/posts/2022/02/02/floss-security).
In reality, distro security depends on a number of factors, such as project activity, developer experience, the level of rigor applied to code reviews, and how often attention is given to specific parts of the codebase that may go untouched for years.
### Missing Security Features
At the moment, desktop Linux [falls behind alternatives](https://discussion.fedoraproject.org/t/fedora-strategy-2028-proposal-fedora-linux-is-as-secure-as-macos/46899/9) like macOS or Android when it comes to certain security features. We hope to see improvements in these areas in the future.
- **Verified boot** on Linux is not as robust as alternatives such as Apple’s [Secure Boot](https://support.apple.com/guide/security/secac71d5623/web) or Android’s [Verified Boot](https://source.android.com/security/verifiedboot). Verified boot prevents persistent tampering by malware and [evil maid attacks](https://en.wikipedia.org/wiki/Evil_Maid_attack), but is still largely [unavailable on even the most advanced distributions](https://discussion.fedoraproject.org/t/has-silverblue-achieved-verified-boot/27251/3).
- **Strong sandboxing** for apps on Linux is severely lacking, even with containerized apps like Flatpaks or sandboxing solutions like Firejail. Flatpak is the most promising sandboxing utility for Linux thus far, but is still deficient in many areas and allows for [unsafe defaults](https://flatkill.org/2020) which permit most apps to trivially bypass their sandbox.
Additionally, Linux falls behind in implementing [exploit mitigations](https://madaidans-insecurities.github.io/linux.html#exploit-mitigations) which are now standard on other operating systems, such as Arbitrary Code Guard on Windows or Hardened Runtime on macOS. Also, most Linux programs and Linux itself are coded in memory-unsafe languages. Memory corruption bugs are responsible for the [majority of vulnerabilities](https://msrc.microsoft.com/blog/2019/07/a-proactive-approach-to-more-secure-code) fixed and assigned a CVE. While this is also true for Windows and macOS, they are quickly making progress on adopting memory-safe languages such as Rust and Swift, respectively.
## Choosing your distribution
Not all Linux distributions are created equal. Our [Linux recommendation page](../desktop.md) is not meant to be an authoritative source on which distribution you should use, but our recommendations *are* aligned with the following guidelines. These are a few things you should keep in mind when choosing a distribution:
### Release cycle
We highly recommend that you choose distributions which stay close to the stable upstream software releases, often referred to as rolling release distributions. This is because frozen release cycle distributions often don’t update package versions and fall behind on security updates.
For frozen distributions such as [Debian](https://debian.org/security/faq#handling), package maintainers are expected to backport patches to fix vulnerabilities rather than bump the software to the “next version” released by the upstream developer. Some security fixes (particularly for less popular software) [do not](https://arxiv.org/abs/2105.14565) receive a [CVE ID](https://en.wikipedia.org/wiki/Common_Vulnerabilities_and_Exposures) at all and therefore do not make it into the distribution with this patching model. As a result, minor security fixes are sometimes held back until the next major release.
We don’t believe holding packages back and applying interim patches is a good idea, as it diverges from the way the developer might have intended the software to work. [Richard Brown](https://rootco.de/aboutme) has a presentation about this:
- [Regular Releases are Wrong, Roll for your life](https://youtu.be/i8c0mg_mS7U) <small>(YouTube)</small>
### Traditional vs Atomic Updates
Traditionally, Linux distributions update by sequentially updating the desired packages. Traditional updates such as those used in Fedora, Arch Linux, and Debian-based distributions can be less reliable if an error occurs while updating.
Distros which use atomic updates, on the other hand, apply updates in full or not at all. On an atomic distribution, if an error occurs while updating (perhaps due to a power failure), nothing is changed on the system.
The atomic update method can achieve reliability with this model and is used for [distributions](../desktop.md#atomic-distributions) like Silverblue and NixOS. [Adam Šamalík](https://twitter.com/adsamalik) provides a presentation on how `rpm-ostree` works with Silverblue:
- [Let's try Fedora Silverblue — an immutable desktop OS! - Adam Šamalík](https://youtu.be/-hpV5l-gJnQ) <small>(YouTube)</small>
### “Security-focused” distributions
There is often some confusion between “security-focused” distributions and “pentesting” distributions. A quick search for “the most secure Linux distribution” will often give results like Kali Linux, Black Arch, or Parrot OS. These distributions are offensive penetration testing distributions that bundle tools for testing other systems. They don’t include any “extra security” or defensive mitigations intended for regular use.
### Arch-based distributions
Arch and Arch-based distributions are not recommended for those new to Linux (regardless of distribution) as they require regular [system maintenance](https://wiki.archlinux.org/title/System_maintenance). Arch does not have a distribution update mechanism for the underlying software choices. As a result you have to stay aware with current trends and adopt technologies on your own as they supersede older practices.
For a secure system, you are also expected to have sufficient Linux knowledge to properly set up security for their system such as adopting a [mandatory access control](#mandatory-access-control) system, setting up [kernel module](https://en.wikipedia.org/wiki/Loadable_kernel_module#Security) blacklists, hardening boot parameters, manipulating [sysctl](https://en.wikipedia.org/wiki/Sysctl) parameters, and knowing what components they need such as [Polkit](https://en.wikipedia.org/wiki/Polkit).
Anyone using the [Arch User Repository (AUR)](https://wiki.archlinux.org/title/Arch_User_Repository) **must** be comfortable auditing PKGBUILDs that they download from that service. AUR packages are community-produced content and are not vetted in any way, and therefore are vulnerable to software [:material-package-variant-closed-remove: Supply Chain Attacks](../basics/common-threats.md#attacks-against-certain-organizations){ .pg-viridian }, which has in fact happened [in the past](https://bleepingcomputer.com/news/security/malware-found-in-arch-linux-aur-package-repository).
The AUR should always be used sparingly, and often there is a lot of bad advice on various pages which direct people to blindly use [AUR helpers](https://wiki.archlinux.org/title/AUR_helpers) without sufficient warning. Similar warnings apply to the use of third-party Personal Package Archives (PPAs) on Debian-based distributions or Community Projects (COPR) on Fedora.
If you are experienced with Linux and wish to use an Arch-based distribution, we generally recommend mainline Arch Linux over any of its derivatives.
Additionally, we recommend **against** these two Arch derivatives specifically:
- **Manjaro**: This distribution holds packages back for 2 weeks to make sure that their own changes don’t break, not to make sure that upstream is stable. When AUR packages are used, they are often built against the latest [libraries](https://en.wikipedia.org/wiki/Library_(computing)) from Arch’s repositories.
- **Garuda**: They use [Chaotic-AUR](https://aur.chaotic.cx) which automatically and blindly compiles packages from the AUR. There is no verification process to make sure that the AUR packages don’t suffer from supply chain attacks.
### Linux-libre kernel and “Libre” distributions
We recommend **against** using the Linux-libre kernel, since it [removes security mitigations](https://phoronix.com/news/GNU-Linux-Libre-5.7-Released) and [suppresses kernel warnings](https://news.ycombinator.com/item?id=29674846) about vulnerable microcode.
### Mandatory access control
Mandatory access control is a set of additional security controls which help to confine parts of the system such as apps and system services. The two common forms of mandatory access control found in Linux distributions are [SELinux](https://github.com/SELinuxProject) and [AppArmor](https://apparmor.net). Fedora and Tumbleweed use SELinux by default, with Tumbleweed offering an option in its installer to choose AppArmor instead.
SELinux on [Fedora](https://docs.fedoraproject.org/en-US/quick-docs/selinux-getting-started) confines Linux containers, virtual machines, and service daemons by default. AppArmor is used by the snap daemon for [sandboxing](https://snapcraft.io/docs/security-sandboxing) snaps which have [strict](https://snapcraft.io/docs/snap-confinement) confinement such as [Firefox](https://snapcraft.io/firefox). There is a community effort to confine more parts of the system in Fedora with the [ConfinedUsers](https://fedoraproject.org/wiki/SIGs/ConfinedUsers) special interest group.
## General Recommendations
### Drive Encryption
Most Linux distributions have an option within its installer for enabling [LUKS](../encryption.md#linux-unified-key-setup) FDE. If this option isn’t set at installation time, you will have to back up your data and re-install, as encryption is applied after [disk partitioning](https://en.wikipedia.org/wiki/Disk_partitioning), but before [file systems](https://en.wikipedia.org/wiki/File_system) are formatted. We also suggest securely erasing your storage device:
- [Secure Data Erasure :material-arrow-right-drop-circle:](https://blog.privacyguides.org/2022/05/25/secure-data-erasure)
### Swap
Consider using [ZRAM](https://wiki.archlinux.org/title/Zram#Using_zram-generator) instead of a traditional swap file or partition to avoid writing potentially sensitive memory data to persistent storage (and improve performance). Fedora-based distributions [use ZRAM by default](https://fedoraproject.org/wiki/Changes/SwapOnZRAM).
If you require suspend-to-disk (hibernation) functionality, you will still need to use a traditional swap file or partition. Make sure that any swap space you do have on a persistent storage device is [encrypted](https://wiki.archlinux.org/title/Dm-crypt/Swap_encryption) at a minimum to mitigate some of these threats.
### Proprietary Firmware (Microcode Updates)
Some Linux distributions (such as [Linux-libre](https://en.wikipedia.org/wiki/Linux-libre)-based or DIY distros) don’t come with the proprietary [microcode](https://en.wikipedia.org/wiki/Microcode) updates which patch critical security vulnerabilities. Some notable examples of these vulnerabilities include [Spectre](https://en.wikipedia.org/wiki/Spectre_(security_vulnerability)), [Meltdown](https://en.wikipedia.org/wiki/Meltdown_(security_vulnerability)), [SSB](https://en.wikipedia.org/wiki/Speculative_Store_Bypass), [Foreshadow](https://en.wikipedia.org/wiki/Foreshadow), [MDS](https://en.wikipedia.org/wiki/Microarchitectural_Data_Sampling), [SWAPGS](https://en.wikipedia.org/wiki/SWAPGS_(security_vulnerability)), and other [hardware vulnerabilities](https://kernel.org/doc/html/latest/admin-guide/hw-vuln/index.html).
We **highly recommend** that you install microcode updates, as they contain important security patches for the CPU which can not be fully mitigated in software alone. Fedora and openSUSE both apply microcode updates by default.
### Updates
Most Linux distributions will automatically install updates or remind you to do so. It is important to keep your OS up to date so that your software is patched when a vulnerability is found.
Some distributions (particularly those aimed at advanced users) are more bare bones and expect you to do things yourself (e.g. Arch or Debian). These will require running the "package manager" (`apt`, `pacman`, `dnf`, etc.) manually in order to receive important security updates.
Additionally, some distributions will not download firmware updates automatically. For that, you will need to install [`fwupd`](https://wiki.archlinux.org/title/Fwupd).
### Permission Controls
Desktop environments that support the [Wayland](https://wayland.freedesktop.org) display protocol are [more secure](https://lwn.net/Articles/589147) than those that only support X11. Moreover, we *generally* recommend installing and using applications which are sandboxed such as those obtained via **Flatpak**. Flatpak supports the [`security-context-v1`](https://github.com/flatpak/flatpak/pull/4920) protocol and the ability to filter D-Bus protocols, which allow Flatpak to properly identify apps for the purpose of sandboxing them through permission controls.[^1] Conversely, applications outside sandboxes are free to perform privileged actions such as capturing your screen, either by [overwriting the portal permission store](https://invent.kde.org/plasma/xdg-desktop-portal-kde/-/issues/7#note_1112260), or [making use of privileged Wayland protocols](https://github.com/swaywm/sway/pull/7648#issuecomment-2507730794).
[^1]: This exposes a reliable way for Wayland compositors to get identifying information about a client. Compositors can then apply security policies if desirable. [https://github.com/flatpak/flatpak/commit/f0e626a4b60439f211f06d35df74b675a9ef42f4](https://github.com/flatpak/flatpak/commit/f0e626a4b60439f211f06d35df74b675a9ef42f4)
## Privacy Tweaks
### MAC Address Randomization
Many desktop Linux distributions (Fedora, openSUSE, etc.) come with [NetworkManager](https://en.wikipedia.org/wiki/NetworkManager) to configure Ethernet and Wi-Fi settings.
It is possible to randomize the [MAC address](https://en.wikipedia.org/wiki/MAC_address) when using NetworkManager. This provides a bit more privacy on Wi-Fi networks as it makes it harder to track specific devices on the network you’re connected to. It does [**not**](https://papers.mathyvanhoef.com/wisec2016.pdf) make you anonymous.
In the terminal, create a new file `/etc/NetworkManager/conf.d/00-macrandomize.conf` and add the following to it:
```text
[device]
wifi.scan-rand-mac-address=yes
[connection]
wifi.cloned-mac-address=random
ethernet.cloned-mac-address=random
```
Then, restart NetworkManager:
```sh
systemctl restart NetworkManager
```
Optionally, changing the connection parameter from `random` to `stable` will give you a random MAC address *per network*, but keep it stable for that network when you reconnect to it later. Using `random` will give you a random MAC address *per connection*. This may be desirable for networks with captive portals or where you have a static DHCP assignment, at the expense of making you more identifiable by a single network operator you connect to multiple times.
If you are using [systemd-networkd](https://en.wikipedia.org/wiki/Systemd#Ancillary_components), you will need to set [`MACAddressPolicy=random`](https://freedesktop.org/software/systemd/man/systemd.link.html#MACAddressPolicy=) which will enable [RFC 7844 (Anonymity Profiles for DHCP Clients)](https://freedesktop.org/software/systemd/man/systemd.network.html#Anonymize=).
MAC address randomization is primarily beneficial for Wi-Fi connections. For Ethernet connections, randomizing your MAC address provides little (if any) benefit, because a network administrator can trivially identify your device by other means (such as inspecting the port you are connected to on the network switch). Randomizing Wi-Fi MAC addresses depends on support from the Wi-Fi’s firmware.
### Other Identifiers
There are other system identifiers which you may wish to be careful about. You should give this some thought to see if it applies to your [threat model](../basics/threat-modeling.md):
- **Hostnames:** Your system's hostname is shared with the networks you connect to. You should avoid including identifying terms like your name or operating system in your hostname, instead sticking to generic terms or random strings.
- **Usernames:** Similarly, your username is used in a variety of ways across your system. Consider using generic terms like "user" rather than your actual name.
### System Counting
The Fedora Project [counts](https://fedoraproject.org/wiki/Changes/DNF_Better_Counting) how many unique systems access its mirrors by using a [`countme`](https://fedoraproject.org/wiki/Changes/DNF_Better_Counting#Detailed_Description) variable instead of a unique ID. Fedora does this to determine load and provision better servers for updates where necessary.
This [option](https://dnf.readthedocs.io/en/latest/conf_ref.html#options-for-both-main-and-repo) is currently off by default. We recommend adding `countme=false` to `/etc/dnf/dnf.conf` just in case it is enabled in the future. On systems that use `rpm-ostree` such as Silverblue, the `countme` option is disabled by masking the [rpm-ostree-countme](https://fedoramagazine.org/getting-better-at-counting-rpm-ostree-based-systems) timer.
openSUSE also uses a [unique ID](https://en.opensuse.org/openSUSE:Statistics) to count systems, which can be disabled by emptying the `/var/lib/zypp/AnonymousUniqueId` file.
content/wiki/os/macos-overview.md
+311
View File
@@ -0,0 +1,311 @@
---
title: macOS Overview
icon: material/apple-finder
description: macOS is Apple's desktop operating system that works with their hardware to provide strong security.
---
**macOS** is a Unix operating system developed by Apple for their Mac computers. To enhance privacy on macOS, you can disable telemetry features and harden existing privacy and security settings.
Older Intel-based Macs and Hackintoshes do not support all the security features that macOS offers. To enhance data security, we recommend using a newer Mac with [Apple Silicon](https://support.apple.com/HT211814).
## Privacy Notes
There are a few notable privacy concerns with macOS that you should consider. These pertain to the operating system itself, and not Apple's other apps and services.
### Activation Lock
Brand-new Apple Silicon devices can be set up without an internet connection. However, recovering or resetting your Mac will **require** an internet connection to Apple's servers to check against the Activation Lock database of lost or stolen devices.
### App Revocation Checks
macOS performs online checks when you open an app to verify whether an app contains known malware, and whether the developer’s signing certificate is revoked.
Apple's OCSP service uses HTTPS encryption, so only they are able to see which apps you open. They've [posted information](https://support.apple.com/HT202491) about their logging policy for this service. They additionally [promised](http://lapcatsoftware.com/articles/2024/8/3.html) to add a mechanism for people to opt-out of this online check, but this has not been added to macOS.
While you [can](https://eclecticlight.co/2021/02/23/how-to-run-apps-in-private) manually opt out of this check relatively easily, we recommend against doing so unless you would be badly compromised by the revocation checks performed by macOS, because they serve an important role in ensuring compromised apps are blocked from running.
## Recommended Configuration
Your account when you first set up your Mac will be an Administrator account, which has higher privileges than a Standard user account. macOS has a number of protections which prevent malware and other programs from abusing your Administrator privileges, so it is generally safe to use this account.
However, exploits in protective utilities like `sudo` have been [discovered in the past](https://bogner.sh/2014/03/another-mac-os-x-sudo-password-bypass). If you want to avoid the possibility that programs you run abuse your Administrator privileges, you could consider creating a second, Standard user account which you use for day-to-day operations. This has the added benefit of making it more obvious when an app needs admin access, because it will prompt you for credentials every time.
If you do use a second account, it is not strictly required to ever log in to your original Administrator account from the macOS login screen. When you are doing something as a Standard user which requires Administrator permissions, the system should prompt you for authentication, where you can enter your Administrator credentials as your Standard user on a one-time basis. Apple provides [guidance](https://support.apple.com/HT203998) on hiding your Administrator account if you prefer to only see a single account on your login screen.
### iCloud
When you use Apple services like iCloud, most of your information is stored on their servers and secured with keys *which Apple has access to* by default. This is called [Standard Data Protection](https://support.apple.com/en-us/102651) by Apple.
Therefore, if you use iCloud you should [enable **Advanced Data Protection**](https://support.apple.com/HT212520). This encrypts nearly all of your iCloud data with keys stored on your devices (end-to-end encryption), rather than Apple's servers, so that your iCloud data is secured in the event of a data breach, and otherwise hidden from Apple.
If you want to be able to install apps from the App Store but don't want to enable iCloud, you can sign in to your Apple Account from the App Store instead of **System Settings**.
### System Settings
There are a number of built-in settings you should confirm or change to harden your system. Open the **Settings** app:
#### Bluetooth
- [ ] Turn off **Bluetooth** (unless you are currently using it)
#### Network
Depending on if you are using **Wi-Fi** or **Ethernet** (denoted by a green dot and the word "connected"), click on the corresponding icon.
Click on the "Details" button by your network name:
- [x] Select **Rotating** under **Private Wi-Fi address**
- [x] Turn on **Limit IP address tracking**
##### Firewall
Your firewall blocks unwanted network connections. The stricter your firewall settings are, the more secure your Mac is. However, certain services will be blocked. You should configure your firewall to be as strict as you can without blocking services you use.
- [x] Turn on **Firewall**
Click the **Options** button:
- [x] Turn on **Block all incoming connections**
If this configuration is too strict, you can come back and uncheck this. However, macOS will typically prompt you to allow incoming connections for an app if the app requests it.
#### General
By default, your device name will be something like "[your name]'s iMac". Because this name is [publicly broadcast on your network](https://support.apple.com/guide/mac-help/change-computers-local-hostname-mac-mchlp2322/26/mac/26#:~:text=The%20local%20hostname%2C%20or%20local%20network%20name%2C%20is%20displayed%20at%20the%20bottom%20of%20the%20Sharing%20settings%20window.%20It%20identifies%20your%20Mac%20to%20Bonjour%2Dcompatible%20services.), you'll want to change your device name to something generic like "Mac".
Click on **About** and type your desired device name into the **Name** field.
##### Software Updates
You should automatically install all available updates to make sure your Mac has the latest security fixes.
Click the small :material-information-outline: icon next to **Automatic Updates**:
- [x] Turn on **Download new updates when available**
- [x] Turn on **Install macOS updates**
- [x] Turn on **Install Security Responses and system files**
#### Apple Intelligence & Siri
If you do not use these features on macOS, you should disable them:
- [ ] Turn off **Apple Intelligence**
- [ ] Turn off **Siri**
**[Apple Intelligence](https://apple.com/legal/privacy/data/en/intelligence-engine)** is only available if your device supports it. Apple Intelligence uses a combination of on-device processing and their [Private Cloud Compute](https://security.apple.com/blog/private-cloud-compute) for things that take more processing power than your device can provide.
To see a report of all the data sent via Apple Intelligence, you can navigate to **Privacy & Security** → **Apple Intelligence Report** and press **Export Activity** to see activity from the either the last 15 minutes or 7 days, depending on what you set it for. Similar to the **App Privacy Report** which shows you the recent permissions accessed by the apps on your phone, the Apple Intelligence Report likewise shows what is being sent to Apple's servers while using Apple Intelligence.
By default, ChatGPT integration is disabled. If you don't want ChatGPT integration anymore, you can navigate to **ChatGPT**:
- [ ] Turn off **Use ChatGPT**
You can also have it ask for confirmation every time if you leave ChatGPT integration on:
- [x] Turn on **Confirm Requests**
<div class="admonition warning" markdown>
<p class="admonition-title">Warning</p>
Any request made with ChatGPT will be sent to ChatGPT's servers, there is no on-device processing and no PCC like with Apple Intelligence.
</div>
#### Privacy & Security
Whenever an application requests a permission, it will show up here. You can decide which applications you want to allow or deny specific permissions.
##### Location Services
You can individually allow location services per-app. If you don't need apps to use your location, turning off location services entirely is the most private option.
- [ ] Turn off **Location Services**
##### Analytics & Improvements
Decide whether you want to share analytics data with Apple and app developers.
##### Apple Advertising
Decide whether you want personalized ads based on your usage.
- [ ] Turn off **Personalized Ads**
##### FileVault
On modern devices with a Secure Enclave (Apple T2 Security Chip, Apple Silicon), your data is always encrypted, but is decrypted automatically by a hardware key if your device doesn't detect it's been tampered with. Enabling [FileVault](../encryption.md#filevault) additionally requires your password to decrypt your data, greatly improving security, especially when powered off or before the first login after powering on.
On older Intel-based Mac computers, FileVault is the only form of disk encryption available by default, and should always be enabled.
- [x] Click **Turn On**
##### Lockdown Mode
**[Lockdown Mode](https://support.apple.com/guide/mac-help/lock-mac-targeted-a-cyberattack-ibrw66f4e191/mac)** disables some features in order to improve security. Some apps or features won't work the same way they do when it's off. For example, Javascript Just-In-Time ([JIT](https://hacks.mozilla.org/2017/02/a-crash-course-in-just-in-time-jit-compilers)) compilation and [WebAssembly](https://developer.mozilla.org/docs/WebAssembly) are disabled in Safari with Lockdown Mode enabled. We recommend enabling Lockdown Mode and seeing whether it significantly impacts daily usage.
- [x] Click **Turn On**
### MAC Address Randomization
macOS uses a randomized MAC address when [performing Wi-Fi scans](https://support.apple.com/guide/security/privacy-features-connecting-wireless-networks-secb9cb3140c/web) while disconnected from a network.
You can set your [MAC address to be randomized](https://support.apple.com/en-us/102509) per network and rotate occasionally to prevent tracking between networks and on the same network over time.
Go to **System Settings** → **Network** → **Wi-Fi** → **Details** and set **Private Wi-Fi address** to either **Fixed** if you want a fixed but unique address for the network you're connected to, or **Rotating** if you want it to change over time.
Consider changing your hostname as well, which is another device identifier that's broadcast on the network you're connected to. You may wish to set your hostname to something generic like "MacBook Air", "Laptop", "John's MacBook Pro", or "iPhone" in **System Settings** → **General** → **Sharing**.
## Security Protections
macOS employs defense in depth by relying on multiple layers of software and hardware-based protections, with different properties. This ensures that a failure in one layer does not compromise the system's overall security.
### Software Security
<div class="admonition warning" markdown>
<p class="admonition-title">Warning</p>
macOS allows you to install beta updates. These are unstable and may come with [extra telemetry](https://beta.apple.com/privacy) since they're for testing purposes. Because of this, we recommend you avoid beta software in general.
</div>
#### Signed System Volume
macOS's system components are protected in a read-only [signed system volume](https://support.apple.com/guide/security/signed-system-volume-security-secd698747c9/web), meaning that neither you nor malware can alter important system files.
The system volume is verified while it's running and any data that's not signed with a valid cryptographic signature from Apple will be rejected.
#### System Integrity Protection
macOS sets certain security restrictions that can't be overridden. These are called [Mandatory Access Controls](https://support.apple.com/guide/security/system-integrity-protection-secb7ea06b49/1/web/1), and they form the basis of the sandbox, parental controls, and [System Integrity Protection](https://support.apple.com/en-us/102149) on macOS.
System Integrity Protection makes critical file locations read-only to protect against modification from malicious code. This is on top of the hardware-based Kernel Integrity Protection that keeps the kernel from being modified in-memory.
#### Application Security
##### App Sandbox
On macOS, whether an app is sandboxed is determined by the developer when they sign it. The [App Sandbox](https://developer.apple.com/documentation/xcode/configuring-the-macos-app-sandbox) protects against vulnerabilities in the apps you run by limiting what a malicious actor can access in the event that the app is exploited. The App Sandbox *alone* can't protect against [:material-package-variant-closed-remove: Supply Chain Attacks](../basics/common-threats.md#attacks-against-certain-organizations){ .pg-viridian } by malicious developers. For that, sandboxing needs to be enforced by someone other than the developer themselves, as it is on the [App Store](https://support.apple.com/guide/security/gatekeeper-and-runtime-protection-sec5599b66df/1/web/1#:~:text=All%20apps%20from%20the%20App%20Store%20are%20sandboxed%20to%20restrict%20access%20to%20data%20stored%20by%20other%20apps.).
<div class="admonition warning" markdown>
<p class="admonition-title">Warning</p>
Software downloaded from outside the official App Store is not required to be sandboxed. If your threat model prioritizes defending against [:material-bug-outline: Passive Attacks](../basics/common-threats.md#security-and-privacy){ .pg-orange }, then you may want to check if the software you download outside the App Store is sandboxed, which is up to the developer to *opt in*.
</div>
You can check if an app uses the App Sandbox in a few ways:
You can check if apps that are already running are sandboxed using the [Activity Monitor](https://developer.apple.com/documentation/security/protecting-user-data-with-app-sandbox#Verify-that-your-app-uses-App-Sandbox).
<div class="admonition warning" markdown>
<p class="admonition-title">Warning</p>
Just because one of an app's processes is sandboxed doesn't mean they all are.
</div>
Alternatively, you can check apps before you run them by running this command in the terminal:
``` zsh
codesign -dvvv --entitlements - <path to your app>
```
If an app is sandboxed, you should see the following output:
``` zsh
[Key] com.apple.security.app-sandbox
[Value]
[Bool] true
```
If you find that the app you want to run is not sandboxed, then you may employ methods of [compartmentalization](../basics/common-threats.md#security-and-privacy) such as virtual machines or separate devices, use a similar app that is sandboxed, or choose to not use the non-sandboxed app altogether.
##### Hardened Runtime
The [Hardened Runtime](https://developer.apple.com/documentation/security/hardened_runtime) is an extra form of protection for apps that prevents certain classes of exploits. It improves the security of apps against exploitation by disabling certain features like JIT.
You can check if an app uses the Hardened Runtime using this command:
``` zsh
codesign -dv <path to your app>
```
If Hardened Runtime is enabled, you will see `flags=0x10000(runtime)`. The `runtime` output means Hardened Runtime is enabled. There might be other flags, but the runtime flag is what we're looking for here.
You can enable a column in Activity Monitor called "Restricted" which is a flag that prevents programs from injecting code via macOS's [dynamic linker](https://pewpewthespells.com/blog/blocking_code_injection_on_ios_and_os_x.html). Ideally, this should say "Yes".
##### Antivirus
macOS comes with two forms of [malware defense](https://support.apple.com/guide/security/protecting-against-malware-sec469d47bd8/1/web/1):
1. Protection against launching malware in the first place is provided by the App Store's review process for App Store applications, or *Notarization* (part of *Gatekeeper*), a process where third-party apps are scanned for known malware by Apple before they are allowed to run. Apps are required to be signed by the developers using a key given to them by Apple. This ensures that you are running software from the real developers. Notarization also requires that developers enable the Hardened Runtime for their apps, which limits methods of exploitation.
2. Protection against other malware and remediation from existing malware on your system is provided by *XProtect*, a more traditional antivirus software built-in to macOS.
We recommend against installing third-party antivirus software as they typically do not have the system-level access required to properly function anyway, because of Apple's limitations on third-party apps, and because granting the high levels of access they do ask for often poses an even greater security and privacy risk to your computer.
##### Backups
macOS comes with automatic backup software called [Time Machine](https://support.apple.com/HT201250), so you can create [encrypted backups](https://support.apple.com/guide/mac-help/keep-your-time-machine-backup-disk-secure-mh21241/mac) to an external drive or a network drive in the event of corrupted/deleted files.
### Hardware Security
Many modern security features in macOS—such as modern Secure Boot, hardware-level exploit mitigation, OS integrity checks, and file-based encryption—rely on Apple Silicon, and Apple's newer hardware always has the [best security](https://support.apple.com/guide/security/apple-soc-security-sec87716a080/1/web/1). We only encourage the use of Apple Silicon, and not older Intel-based Mac computers or Hackintoshes.
Some of these modern security features are available on older Intel-based Mac computers with the Apple T2 Security Chip, but that chip is susceptible to the *checkm8* exploit which could compromise its security.
If you use Bluetooth accessories such as a keyboard, we recommend that you use official Apple ones as their firmware will [automatically be updated](https://support.apple.com/en-us/120303#:~:text=Firmware%20updates%20are%20automatically%20delivered%20in%20the%20background%20while%20the%20Magic%20Keyboard%20is%20actively%20paired%20to%20a%20device%20running%20macOS%2C%20iOS%2C%20iPadOS%2C%20or%20tvOS.) for you by macOS. Using third party accessories is fine, but you should remember to install firmware updates for them regularly.
Apple's SoCs focus on [minimizing attack surface](https://support.apple.com/en-vn/guide/security/secf020d1074/web#:~:text=Security%2Dfocused%20hardware%20follows%20the%20principle%20of%20supporting%20limited%20and%20discretely%20defined%20functions%20to%20minimize%20attack%20surface.) by relegating security functions to dedicated hardware with limited functionality.
#### Boot ROM
macOS prevents malware persistence by only allowing official Apple software to run at boot time; this is known as [secure boot](https://support.apple.com/en-vn/guide/security/secac71d5623/1/web/1). Mac computers verify this with a bit of read-only memory on the SoC called the [boot ROM](https://support.apple.com/en-vn/guide/security/aside/sec5240db956/1/web/1), which is [laid down during the manufacturing of the chip](https://support.apple.com/en-vn/guide/security/secf020d1074/1/web/1#:~:text=which%20is%20laid%20down%20during%20Apple%20SoC%20fabrication).
The boot ROM forms the hardware root of trust. This ensures that malware cannot tamper with the boot process, since the boot ROM is immutable. When your Mac boots up, the boot ROM is the first thing that runs, forming the first link in the chain of trust.
Mac computers can be configured to boot in [three security modes](https://support.apple.com/guide/deployment/startup-security-dep5810e849c/web#dep32fb404e1): *Full Security*, *Reduced Security*, and *Permissive Security*, with the default setting being Full Security. You should ideally be using Full Security mode and avoid things like **[kernel extensions](https://support.apple.com/guide/deployment/system-extensions-in-macos-depa5fb8376f/web#dep51e097f45)** that force you to lower your security mode. Make sure to [check](https://support.apple.com/guide/mac-help/change-security-settings-startup-disk-a-mac-mchl768f7291/mac) that you're using Full Security mode.
#### Secure Enclave
The **[Secure Enclave](https://support.apple.com/guide/security/secure-enclave-sec59b0b31ff/web)** is a security chip built into devices with Apple Silicon which is responsible for storing and generating encryption keys for data at rest as well as Face ID and Touch ID data. It contains its own [separate boot ROM](https://support.apple.com/en-vn/guide/security/sec59b0b31ff/web#sec43006c49f).
You can think of the Secure Enclave as your device's security hub: it has an AES encryption engine and a mechanism to securely store your encryption keys, and it's separated from the rest of the system, so even if the main processor is compromised, it should still be safe.
#### Touch ID
Apple's Touch ID feature allows you to securely unlock your devices using biometrics.
Your biometric data [never leaves your device](https://www.apple.com/legal/privacy/data/en/touch-id/#:~:text=Touch%C2%A0ID%20data%20does%20not%20leave%20your%20device%2C%20and%20is%20never%20backed%20up%20to%20iCloud%20or%20anywhere%20else.); it's stored only in the Secure Enclave.
#### Hardware Microphone Disconnect
All laptops with Apple Silicon or the T2 chip feature a [hardware disconnect](https://support.apple.com/guide/security/hardware-microphone-disconnect-secbbd20b00b/web) for the built-in microphone whenever the lid is closed. This means that there is no way for an attacker to listen to your Mac's microphone even if the operating system is compromised.
Note that the camera does not have a hardware disconnect, since its view is obscured when the lid is closed anyway.
#### Secure Camera Indicator
The built-in camera in a Mac is designed so that the camera can't turn on without the camera indicator light [also turning on](https://support.apple.com/en-us/102177#:~:text=The%20camera%20is%20engineered%20so%20that%20it%20can’t%20activate%20without%20the%20camera%20indicator%20light%20also%20turning%20on.%20This%20is%20how%20you%20can%20tell%20if%20your%20camera%20is%20on.).
#### Peripheral Processor Security
Computers have [built-in processors](https://support.apple.com/en-vn/guide/security/seca500d4f2b/1/web/1) other than the main CPU that handle things like networking, graphics, power management, etc. These processors can have insufficient security and become compromised, therefore Apple tries to minimize the need for these processors in their hardware.
When it is necessary to use one of these processors, Apple works with the vendor to ensure that the processor
- runs verified firmware from the primary CPU on startup
- has its own Secure Boot chain
- follows minimum cryptographic standards
- ensures known bad firmware is properly revoked
- has its debug interfaces disabled
- is signed with Apple's cryptographic keys
#### Direct Memory Access Protections
Apple Silicon separates each component that requires [direct memory access](https://support.apple.com/guide/security/direct-memory-access-protections-seca4960c2b5/1/web/1). For example, a Thunderbolt port can't access memory designated for the kernel.
#### Terminal Secure Keyboard Entry
Enable [Secure Keyboard Entry](https://support.apple.com/guide/terminal/use-secure-keyboard-entry-trml109/mac) to prevent other apps from detecting what you type in the terminal.
content/wiki/os/qubes-overview.md
+80
View File
@@ -0,0 +1,80 @@
---
title: "Qubes Overview"
icon: simple/qubesos
description: Qubes is an operating system built around isolating apps within *qubes* (formerly "VMs") for heightened security.
---
[**Qubes OS**](../desktop.md#qubes-os) is an open-source operating system which uses the [Xen](https://en.wikipedia.org/wiki/Xen) hypervisor to provide strong security for desktop computing through isolated *qubes*, (which are Virtual Machines). You can assign each *qube* a level of trust based on its purpose. Qubes OS provides security by using isolation. It only permits actions on a per-case basis and therefore is the opposite of [badness enumeration](https://ranum.com/security/computer_security/editorials/dumb).
## How does Qubes OS work?
Qubes uses [compartmentalization](https://qubes-os.org/intro) to keep the system secure. Qubes are created from templates, the defaults being for Fedora, Debian and [Whonix](../desktop.md#whonix). Qubes OS also allows you to create once-use [disposable](https://qubes-os.org/doc/how-to-use-disposables) *qubes*.
<details class="note" markdown>
<summary>The term <em>qubes</em> is gradually being updated to avoid referring to them as "virtual machines".</summary>
Some of the information here and on the Qubes OS documentation may contain conflicting language as the "appVM" term is gradually being changed to "qube". Qubes are not entire virtual machines, but maintain similar functionalities to VMs.
</details>
![Qubes architecture](../assets/img/qubes/qubes-trust-level-architecture.png)
<figcaption>Qubes Architecture, Credit: What is Qubes OS Intro</figcaption>
Each qube has a [colored border](https://qubes-os.org/screenshots) that can help you keep track of the domain in which it runs. You could, for example, use a specific color for your banking browser, while using a different color for a general untrusted browser.
![Colored border](../assets/img/qubes/r4.0-xfce-three-domains-at-work.png)
<figcaption>Qubes window borders, Credit: Qubes Screenshots</figcaption>
## Why Should I use Qubes?
Qubes OS is useful if your [threat model](../basics/threat-modeling.md) requires strong security and isolation, such as if you think you'll be opening untrusted files from untrusted sources. A typical reason for using Qubes OS is to open documents from unknown sources, but the idea is that if a single qube is compromised it won't affect the rest of the system.
Qubes OS utilizes [dom0](https://wiki.xenproject.org/wiki/Dom0) Xen VM for controlling other *qubes* on the host OS, all of which display individual application windows within dom0's desktop environment. There are many uses for this type of architecture. Here are some tasks you can perform. You can see just how much more secure these processes are made by incorporating multiple steps.
### Copying and Pasting Text
You can [copy and paste text](https://qubes-os.org/doc/how-to-copy-and-paste-text) using `qvm-copy-to-vm` or the below instructions:
1. Press **Ctrl+C** to tell the *qube* you're in that you want to copy something.
2. Press **Ctrl+Shift+C** to tell the *qube* to make this buffer available to the global clipboard.
3. Press **Ctrl+Shift+V** in the destination *qube* to make the global clipboard available.
4. Press **Ctrl+V** in the destination *qube* to paste the contents in the buffer.
### File Exchange
To copy and paste files and directories (folders) from one *qube* to another, you can use the option **Copy to Other AppVM...** or **Move to Other AppVM...**. The difference is that the **Move** option will delete the original file. Either option will protect your clipboard from being leaked to any other *qubes*. This is more secure than air-gapped file transfer. An air-gapped computer will still be forced to parse partitions or file systems. That is not required with the inter-qube copy system.
<details class="note" markdown>
<summary>Qubes do not have their own filesystems.</summary>
You can [copy and move files](https://qubes-os.org/doc/how-to-copy-and-move-files) between *qubes*. When doing so the changes aren't immediately made and can be easily undone in case of an accident. When you run a *qube*, it does not have a persistent filesystem. You can create and delete files, but these changes are ephemeral.
</details>
### Inter-VM Interactions
The [qrexec framework](https://qubes-os.org/doc/qrexec) is a core part of Qubes which allows communication between domains. It is built on top of the Xen library *vchan*, which facilitates [isolation through policies](https://qubes-os.org/news/2020/06/22/new-qrexec-policy-system).
## Connecting to Tor via a VPN
We [recommend](../advanced/tor-overview.md) connecting to the Tor network via a [VPN](../vpn.md) provider, and luckily Qubes makes this easy to do with a combination of ProxyVMs and Whonix.
After [creating a new ProxyVM](https://forum.qubes-os.org/t/configuring-a-proxyvm-vpn-gateway/19061) which connects to the VPN of your choice, you can chain your Whonix qubes to that ProxyVM **before** they connect to the Tor network, by setting the NetVM of your Whonix **Gateway** (`sys-whonix`) to the newly-created ProxyVM.
Your qubes should be configured in a manner similar to this:
| Qube name | Qube description | NetVM |
|-----------------|------------------------------------------------------------------------------------------------------------------|-----------------|
| sys-net | *Your default network qube (pre-installed)* | *n/a* |
| sys-firewall | *Your default firewall qube (pre-installed)* | sys-net |
| ==sys-proxyvm== | The VPN ProxyVM you [created](https://forum.qubes-os.org/t/configuring-a-proxyvm-vpn-gateway/19061) | sys-firewall |
| sys-whonix | Your Whonix Gateway VM | ==sys-proxyvm== |
| anon-whonix | Your Whonix Workstation VM | sys-whonix |
## Additional Resources
For additional information we encourage you to consult the extensive Qubes OS documentation pages located on the [Qubes OS Website](https://qubes-os.org/doc). Offline copies can be downloaded from the Qubes OS [documentation repository](https://github.com/QubesOS/qubes-doc).
- [Arguably the world's most secure operating system](https://opentech.fund/news/qubes-os-arguably-the-worlds-most-secure-operating-system-motherboard) (Open Technology Fund)
- [Software compartmentalization vs. physical separation](https://invisiblethingslab.com/resources/2014/Software_compartmentalization_vs_physical_separation.pdf) (J. Rutkowska)
- [Partitioning my digital life into security domains](https://blog.invisiblethings.org/2011/03/13/partitioning-my-digital-life-into.html) (J. Rutkowska)
- [Related Articles](https://qubes-os.org/news/categories/#articles) (Qubes OS)
content/wiki/os/windows/group-policies.md
+134
View File
@@ -0,0 +1,134 @@
---
title: Group Policy Settings
description: A quick guide to configuring Group Policy to make Windows a bit more privacy respecting.
---
Outside modifying the registry itself, the **Local Group Policy Editor** is the most powerful way to change many aspects of your system without installing third-party tools. Changing these settings requires [Pro Edition](index.md#windows-editions) or better.
These settings should be set on a brand-new installation of Windows. Setting them on your existing installation should work, but may introduce unpredictable behavior and is done at your own risk.
All of these settings have an explanation attached to them in the Group Policy editor which explains exactly what they do, usually in great detail. Please pay attention to those descriptions as you make changes, so you know exactly what we are recommending here. We've also explained some of our choices below whenever the explanation included with Windows is inadequate.
## Administrative Templates
You can find these settings by opening `gpedit.msc` and navigating to **Local Computer Policy** > **Computer Configuration** > **Administrative Templates** in the left sidebar. The headers on this page correspond to folders/subfolders within Administrative Templates, and the bullet points correspond to individual policies.
To change any group policy, double click it and select Enabled or Disabled at the top of the window that appears depending on the recommendations below. Some group policies have additional settings that can be configured, and if that's the case the appropriate settings are noted below as well.
### System
#### Device Guard
- Turn On Virtualization Based Security: **Enabled**
- Platform Security Level: **Secure Boot and DMA Protection**
- Secure Launch Configuration: **Enabled**
#### Internet Communication Management
- Turn off Windows Customer Experience Improvement Program: **Enabled**
- Turn off Windows Error Reporting: **Enabled**
- Turn off the Windows Messenger Customer Experience Improvement Program: **Enabled**
Note that disabling the Windows Customer Experience Improvement Program also disables some other tracking features that can be individually controlled with Group Policy as well. We don't list them all here or disable them because this setting covers that.
#### OS Policies
- Allow Clipboard History: **Disabled**
- Allow Clipboard synchronization across devices: **Disabled**
- Enables Activity Feed: **Disabled**
- Allow publishing of User Activities: **Disabled**
- Allow upload of User Activities: **Disabled**
#### User Profiles
- Turn off the advertising ID: **Enabled**
### Windows Components
#### AutoPlay Policies
AutoRun and AutoPlay are features which allow Windows to run a script or perform some other task when a device is connected, sometimes avoiding security measures that involve user consent. This could allow untrusted devices to run malicious code without your knowledge. It's a security best practice to disable these features, and simply open files on your external disks manually.
- Turn off AutoPlay: **Enabled**
- Disallow Autoplay for nonvolume devices: **Enabled**
- Set the default behavior for AutoRun: **Enabled**
- Default AutoRun Behavior: **Do not execute any AutoRun commands**
#### BitLocker Drive Encryption
You may wish to re-encrypt your operating system drive after changing these settings.
- Choose drive encryption method and cipher strength (Windows Vista, Windows Server 2008, Windows 7): **Enabled**
- Select the encryption method: **AES-256**
Setting the cipher strength for the Windows 7 policy still applies that strength to newer versions of Windows.
##### Operating System Drives
- Require additional authentication at startup: **Enabled**
- Allow enhanced PINs for startup: **Enabled**
Despite the names of these policies, this doesn't *require* you to do anything by default, but it will unlock the *option* to have a more complex setup (such as requiring a PIN at startup in addition to the TPM) in the BitLocker setup wizard.
#### Cloud Content
- Turn off cloud optimized content: **Enabled**
- Turn off cloud consumer account state content: **Enabled**
- Do not show Windows tips: **Enabled**
- Turn off Microsoft consumer experiences: **Enabled**
#### Credential User Interface
- Require trusted path for credential entry: **Enabled**
- Prevent the use of security questions for local accounts: **Enabled**
#### Data Collection and Preview Builds
- Allow Diagnostic Data: **Enabled**
- Options: **Send required diagnostic data** (Pro Edition); or
- Options: **Diagnostic data off** (Enterprise or Education Edition)
- Limit Diagnostic Log Collection: **Enabled**
- Limit Dump Collection: **Enabled**
- Limit optional diagnostic data for Desktop Analytics: **Enabled**
- Options: **Disable Desktop Analytics collection**
- Do not show feedback notifications: **Enabled**
#### File Explorer
- Turn off account-based insights, recent, favorite, and recommended files in File Explorer: **Enabled**
#### MDM
- Disable MDM Enrollment: **Enabled**
#### OneDrive
- Save documents to OneDrive by default: **Disabled**
- Prevent OneDrive from generating network traffic until the user signs in to OneDrive: **Enabled**
- Prevent the usage of OneDrive for file storage: **Enabled**
This last setting disables OneDrive on your system; make sure to change it to **Disabled** if you use OneDrive.
#### Push To Install
- Turn off Push To Install service: **Enabled**
#### Search
- Allow Cortana: **Disabled**
- Don't search the web or display web results in Search: **Enabled**
- Set what information is shared in Search: **Enabled**
- Type of information: **Anonymous info**
#### Sync your settings
- Do not sync: **Enabled**
#### Text input
- Improve inking and typing recognition: **Disabled**
#### Windows Error Reporting
- Do not send additional data: **Enabled**
- Consent > Configure Default consent: **Enabled**
- Consent level: **Always ask before sending data**
content/wiki/os/windows/index.md
+63
View File
@@ -0,0 +1,63 @@
---
title: Windows Overview
icon: material/microsoft-windows
description: Microsoft Windows is a common operating system which is extremely non-private out of the box. Our guide covers making some improvements to your computer without replacing your OS.
---
**Microsoft Windows** is a common OS shipped with many PCs by default. The following guides aim to provide some ways to improve privacy and reduce the default telemetry and data stored by disabling some unnecessary features. Over time, Microsoft adds features to the OS which can sometimes rely on cloud-based services. These features will often require certain types of [optional data](https://privacy.microsoft.com/data-collection-windows) that is sometimes sent to remote servers for processing.
One of the newest examples was called **Recall**, a part of the Copilot AI feature set. Recall periodically screenshots anything you've seen on your PC in order to show it to you at a later date. These "helpful" features create considerable metadata which can be forensically analyzed. In most cases browsing history is sufficient and this feature can be safely disabled. The main concerns with Recall was that the data is stored in a local database that is decrypted when your device is powered on, meaning it is an easy target for hackers if the device ever becomes infected with malware. Recall will not redact sensitive information like copied passwords or financial information from the database, but it does protect against making screenshots of any copyrighted content protected by digital rights management (DRM) systems.
Unfortunately, this feature was added without too much thought about the privacy implications of having such a feature enabled by default (which it now [no longer is](https://wired.com/story/microsoft-recall-off-default-security-concerns)). It is not an isolated example, however. Another example was Microsoft automatically [enabling folder backups to OneDrive](https://neowin.net/news/windows-11-is-now-automatically-enabling-onedrive-folder-backup-without-asking-permission) on new Windows 11 installations without asking for permission.
You can enhance your privacy and security on Windows without downloading any third-party tools with these guides:
- Initial Installation (coming soon)
- [Group Policy Settings](group-policies.md)
- Privacy Settings (coming soon)
- Application Sandboxing (coming soon)
- Security Hardening (coming soon)
<div class="admonition example" markdown>
<p class="admonition-title">This section is new</p>
This section is a work in progress, because it takes considerably more time and effort to make a Windows installation more privacy-friendly than other operating systems.
</div>
## Privacy Notes
Microsoft Windows, particularly those versions aimed at consumers like the **Home** version often don't prioritize privacy-friendly features by [default](https://theguardian.com/technology/2015/jul/31/windows-10-microsoft-faces-criticism-over-privacy-default-settings). As a result we often see more [data collection](https://en.wikipedia.org/wiki/Criticism_of_Microsoft#Telemetry_and_data_collection) than necessary, without any real warnings that this is the default behavior. In an attempt to compete with Google in the advertising space, [Cortana](https://en.wikipedia.org/wiki/Cortana_(virtual_assistant)) has included unique identifiers such as an "advertising ID" in order to correlate usage and assist advertisers in targeted advertising. At launch, telemetry could not be disabled in non-enterprise editions of Windows 10. It still cannot be disabled, but Microsoft added the ability to [reduce](https://extremetech.com/computing/243079-upcoming-windows-update-reduces-spying-microsoft-still-mum-data-collects) the data that is sent to them.
With Windows 11 there are a number of restrictions or defaults such as:
- Requiring the use of a Microsoft account instead of a local account.
- Making it more difficult to find local account options for Windows **Pro** and **Enterprise**.
- Enabling all data collection options by default, requiring users to "opt out".
- Heavily integrating Microsoft services like Bing, OneDrive, and Teams in ways which are difficult to remove and presented as the only option to users.
- Setting the default browser always to Edge, or reverting to Edge if it's changed.
- Adding cloud-based AI features to many areas in Windows and various Microsoft Apps.
- Unnecessarily storing sensitive data. Even data which is stored locally and not sent to Microsoft is still a target for hackers or malware on your device.
Microsoft often uses the automatic updates feature to add new functionality to your device and make changes that collect your data and are enabled by default. Some [privacy features](https://blogs.windows.com/windows-insider/2023/11/16/previewing-changes-in-windows-to-comply-with-the-digital-markets-act-in-the-european-economic-area) such as the option to *opt out* of syncing an online Microsoft account with Windows, require you to select a country in the EEA (European Economic Area) during installation. It can be changed to your real country after Windows is installed.
## Windows Editions
Many critical privacy and security features are unfortunately locked away behind higher-cost editions of Windows, instead of being available in Windows **Home**. Some features missing from **Home** include BitLocker Drive Encryption, Hyper-V, and Windows Sandbox. In our Windows guides we will cover how to use all of these features appropriately, so having a premium edition of Windows will be necessary.
Windows **Enterprise** provides the most flexibility when it comes to configuring privacy and security settings built in to Windows. For example, they are the only editions that allow you to enable the highest level of restrictions on data sent to Microsoft via telemetry tools. Unfortunately, Enterprise is not available for retail purchase, so it may not be available to you.
The best version available for *retail* purchase is Windows **Pro** as it has nearly all the features you'll want to use to secure your device, including BitLocker, Hyper-V, etc. The only thing missing is some of the most restrictive limitations on Microsoft's telemetry, unfortunately.
Students and teachers may be able to obtain a Windows **Education** (equivalent to Enterprise) or **Pro Education** license (equivalent to Pro) for free, including on personal devices, from their educational institution. Many schools partner with Microsoft via OnTheHub or Microsoft Azure for Education, so you can check those sites or your school's benefits page to see if you qualify. Whether or not you are able to get these licenses depends entirely on your institution. This may be the best way for many people to obtain an Enterprise-level edition of Windows for personal use. There are no additional privacy or security risks associated with using an Education license compared to the retail versions.
It is not recommended to use third party modified versions of Windows such as Windows AME. Since modified versions of Windows like Windows AME don't receive updates, security features and antivirus definitions in Windows Defender will fall behind the current threat landscape, opening you up to attacks, thus making you even less secure.
## Obtaining Windows
Currently, only Windows 11 license keys are available for purchase, but these keys will work on Windows 10 as well, so you can still purchase a Windows 11 Pro key to activate a Windows 10 install.
The official [Media Creation Tool](https://microsoft.com/software-download/windows11) is the best way to put a Windows installer on a USB flash drive. Third-party tools like Rufus or Etcher may unexpectedly modify the files, which could lead to boot issues or other troubles when installing.
This tool only lets you install a **Home** or **Pro** installation, as there are no publicly available downloads for Windows **Enterprise** edition. If you have an **Enterprise** license key, you can easily upgrade a **Pro** installation. To do this, install Windows **Pro** without entering a license key during setup, then enter your **Enterprise** key in the Settings app after completing the installation. Your **Pro** install will be upgraded to **Enterprise** automatically after entering a valid license key.
If you are installing an **Education** license then you will typically have a private download link that will be provided alongside your license key when you obtain it from your institution's benefits portal.