description: This topic for the IT professional describes the Advanced Security Audit policy setting, Audit Removable Storage, which determines .
description: This topic for the IT professional describes the Advanced Security Audit policy setting, Audit Removable Storage, which determines when there is a read or a write to a removable drive.
ms.assetid: 1746F7B3-8B41-4661-87D8-12F734AFFB26
ms.prod: W10
ms.mktglfcycl: deploy
@ -15,9 +15,9 @@ author: brianlic-msft
- Windows10
This topic for the IT professional describes the Advanced Security Audit policy setting, **Audit Removable Storage**, which determines .
This topic for the IT professional describes the Advanced Security Audit policy setting, **Audit Removable Storage**, which determines when there is a read or a write to a removable drive.
This topic lists new and updated topics in the [Keep Windows 10 secure](index.md) documentation for [Windows 10 and Windows 10 Mobile](../index.md).
## May 2016
|New or changed topic | Description |
|----------------------|-------------|
| [Microsoft Passport errors during PIN creation](microsoft-passport-errors-during-pin-creation.md) | Added errors 0x80090029 and 0x80070057, and merged entries for error 0x801c03ed. |
| [User Account Control Group Policy and registry key settings](user-account-control-group-policy-and-registry-key-settings.md) | Updated for Windows 10 and Windows Server 2016 Technical Preview |
@ -46,11 +46,4 @@ Membership in the local **Administrators** group, or equivalent, is the minimum
3. Verify that the status for the Application Identity service is **Running**.
Starting with Windows 10, the Application Identity service is now a protected process. Because of this, you can no longer manually set the service **Startup type** to **Automatic**.
description: Microsoft Device Guard is a feature set that consists of both hardware and software system integrity hardening features that revolutionize the Windows operating system’s security.
@ -19,7 +18,6 @@ Microsoft Device Guard is a feature set that consists of both hardware and softw
## Introduction to Device Guard
Today’s security threat landscape is more aggressive than ever before. Modern malicious attacks are focused on revenue generation, intellectual property theft, and targeted system degradation, which results in financial loss. Many of these modern attackers are sponsored by nation states with unknown motives and large cyber terrorism budgets. These threats can enter a company through something as simple as an email message and can permanently damage its reputation for securing its software assets, as well as having significant financial impact. Windows10 introduces several new security features that help mitigate a large percentage of today’s known threats.
It is estimated that more than 300,000 new malware variants are discovered daily. Unfortunately, companies currently use an ancient method to discover this infectious software and prevent its use. In fact, current PCs trust everything that runs until malware signatures determine whether a threat exists; then, the antimalware software attempts to clean the PC, often after the malicious software’s effect has already been noticed. This signature-based system focuses on reacting to an infection and ensuring that the particular infection does not happen again. In this model, the system that drives malware detection relies on the discovery of malicious software; only then can a signature be provided to the client to remediate it, which implies that a computer must be infected first. The time between the detection of the malware and a client being issued a signature could mean the difference between losing data and staying safe.
@ -32,15 +30,12 @@ Device Guard's features revolutionize the Windows operating system’s security
## Device Guard overview
Device Guard is a feature set that consists of both hardware and software system integrity hardening features. These features revolutionize the Windows operating system’s security by taking advantage of new virtualization-based security options and the trust-nothing mobile device operating system model. A key feature in this model is called *configurable code integrity*, which allows your organization to choose exactly which software or trusted software publishers are allowed to run code on your client machines—exactly what has made mobile phone security so successful. In addition, Device Guard offers organizations a way to sign existing line-of-business (LOB) applications so that they can trust their own code, without the requirement that the application be repackaged. Also, this same method of signing provides organizations with a way to trust individual third-party applications. Device Guard—with configurable code integrity, Credential Guard, and AppLocker—is the most complete security defense that any Microsoft product has ever been able to offer a Windows client.
Advanced hardware features such as CPU virtualization extensions, IOMMUs, and SLAT, drive these new client security offerings. By integrating these hardware features further into the core operating system, Windows10 leverages them in new ways. For example, the same type 1 hypervisor technology that is used to run virtual machines in Microsoft Hyper-V is used to isolate core Windows services into a virtualization-based, protected container. This is just one example of how Windows10 integrates advanced hardware features deeper into the operating system to offer comprehensive modern security to its users. These hardware features are now available in consumer and enterprise PC markets and are discussed in detail in the [Hardware considerations](#hardware) section.
Advanced hardware features such as CPU virtualization extensions, IOMMUs, and SLAT, drive these new client security offerings. By integrating these hardware features further into the core operating system, Windows10 leverages them in new ways. For example, the same type 1 hypervisor technology that is used to run virtual machines in Microsoft Hyper-V is used to isolate core Windows services into a virtualization-based, protected container. This is just one example of how Windows10 integrates advanced hardware features deeper into the operating system to offer comprehensive modern security to its users. These hardware features are now available in consumer and enterprise PC markets and are discussed in detail in the [Hardware considerations](#hardware-considerations) section.
Along with these new features, some components of Device Guard are existing tools or technologies that have been included in this strategic security offering to provide customers with the most secure Windows operating system possible. Device Guard is intended as a set of client security features to be used in conjunction with the other threat-resistance features available in the Windows operating system, some of which are mentioned in this guide. In addition to an overview of each feature, this guide walks you through the configuration and deployment of them.
### <a href="" id="config-code"></a>
**Configurable code integrity**
The Windows operating system consists of two operating modes: user mode and kernel mode. The base of the operating system runs within the kernel mode, which is where the Windows operating system directly interfaces with hardware resources. User mode is primarily responsible for running applications and brokering information to and from the kernel mode for hardware resource requests. For example, when an application that is running in user mode needs additional memory, the user mode process must request the resources from kernel mode, not directly from RAM.
@ -53,9 +48,7 @@ Historically, most malware has been unsigned. By simply deploying code integrity
The Device Guard core functionality and protection start at the hardware level. Devices that have processors equipped with SLAT technologies and virtualization extensions, such as Intel Virtualization Technology (VT-x) and AMD-V, will be able to take advantage of virtualization-based security (VBS) features that enhance Windows security. Device Guard leverages VBS to isolate core Windows services that are critical to the security and integrity of the operating system. This isolation removes the vulnerability of these services from both the user and kernel modes and acts as an impenetrable barrier for most malware used today. One of these isolated services, called the Windows Code Integrity service, drives the Device Guard kernel mode configurable code integrity feature. This prevents code that has penetrated the kernel mode operations from compromising the code integrity service.
Another Windows 10 feature that employs VBS is Credential Guard. Credential Guard provides additional protection to Active Directory domain users by storing domain credentials within the virtualization container that hosts the Windows security services, such as code integrity. By isolating these domain credentials from the active user mode and kernel mode, they have a much lower risk of being stolen. For more information about how Credential Guard complements Device Guard, see the [Device Guard with Credential Guard](#dg-with-cg) section. For information about how to enable Credential Guard, see the [Enable Credential Guard](#enable-cg) section.
### <a href="" id="dg-with-applocker"></a>
Another Windows 10 feature that employs VBS is Credential Guard. Credential Guard provides additional protection to Active Directory domain users by storing domain credentials within the virtualization container that hosts the Windows security services, such as code integrity. By isolating these domain credentials from the active user mode and kernel mode, they have a much lower risk of being stolen. For more information about how Credential Guard complements Device Guard, see the [Device Guard with Credential Guard](#device-guard-with-credential-guard) section. For information about how to enable Credential Guard, see the [Enable Credential Guard](#enable-credential-guard) section.
**Device Guard with AppLocker**
@ -63,12 +56,8 @@ Although AppLocker is not considered a new Device Guard feature, it complements
**Note**One example in which Device Guard functionality needs AppLocker supplementation is when your organization would like to limit universal applications. Universal applications have already been validated by Microsoft to be trustworthy to run, but an organization may not want to allow specific universal applications to run in their environment. You can accomplish this enforcement by using an AppLocker rule.
AppLocker and Device Guard should run side-by-side in your organization, which offers the best of both security features at the same time and provides the most comprehensive security to as many devices as possible. In addition to these features, Microsoft recommends that you continue to maintain an enterprise antivirus solution for a well-rounded enterprise security portfolio.
### <a href="" id="dg-with-cg"></a>
**Device Guard with Credential Guard**
Although Credential Guard is not a feature within Device Guard, many organizations will likely deploy Credential Guard alongside Device Guard for additional protection against credential theft. Similar to virtualization-based protection of kernel mode code integrity, Credential Guard leverages hypervisor technology to protect domain credentials. This mitigation is targeted at resisting the use of pass-the-hash and pass-the-ticket techniques. By employing multifactor authentication with Credential Guard, organizations can gain additional protection against such threats. For information about how to deploy Credential Guard to your Windows10 Enterprise clients, see the [Enable Credential Guard](#enable-cg) section. In addition to the client-side enablement of Credential Guard, organizations can deploy mitigations at both the CA and domain controller level to help prevent credential theft. Microsoft will be releasing details about these additional mitigations in the future.
@ -86,42 +75,40 @@ You can easily manage Device Guard features by using the familiar enterprise and
-**Windows PowerShell**. Windows PowerShell is primarily used to create and service code integrity policies. These policies represent the most powerful component of Device Guard. For a step-by-step walkthrough of how to create, audit, service, enforce, and deploy code integrity policies, see the [Code integrity policies](#code-integrity-policies) section.
These options provide the same experience you are used to in order to manage your existing enterprise management solutions. For more information about how to manage and deploy Device Guard hardware and code integrity features in your organization, see the [Device Guard deployment](#dg-deployment) section.
## Plan for Device Guard
## Plan for Device Guard
In this section, you will learn about the following topics:
- [Approach enterprise code integrity deployment](#approach-enterprise). Device Guard deployment in your organization requires a planned approach. In this section, you get high-level recommendations for how to approach enterprise code integrity deployment in your organization.
- [Approach enterprise code integrity deployment](#approach-enterprise-code-integrity-deployment). Device Guard deployment in your organization requires a planned approach. In this section, you get high-level recommendations for how to approach enterprise code integrity deployment in your organization.
- [Device Guard deployment scenarios](#device-guard-deployment). When you plan for Device Guard deployment, Microsoft recommends that you categorize each device in your organization into a deployment scenario. These scenarios will provide a roadmap for your Device Guard deployment.
- [Device Guard deployment scenarios](#device-guard-deployment-scenarios). When you plan for Device Guard deployment, Microsoft recommends that you categorize each device in your organization into a deployment scenario. These scenarios will provide a roadmap for your Device Guard deployment.
- [Code signing adoption](#code-signing-adoption). Code signing is important to the security that Device Guard provides. This section outlines the options for code signing and the benefits and disadvantages of each method.
- [Hardware considerations](#hardware). Several Device Guard features require advanced hardware. This section outlines the requirements for each of those features and what to look for during your next hardware refresh.
- [Hardware considerations](#hardware-considerations). Several Device Guard features require advanced hardware. This section outlines the requirements for each of those features and what to look for during your next hardware refresh.
## Approach enterprise code integrity deployment
Enterprises that want to consider Device Guard should not expect deployment to their entire organization overnight. Device Guard implementation requires that you plan for both end-user and IT pro impact. In addition, the deployment of Device Guard features to your enterprise requires a planned, phased approach to ensure that end-user systems are fully capable and ready to enforce these new security restrictions. Perform the following high-level tasks to approach the deployment of Device Guard to your enterprise:
1.**Group devices into similar functions**. Categorize machines into the groups described in the [Device Guard deployment scenarios](#device-guard-deployment) section. This begins the roadmap for your Device Guard deployment and provides groups of easier and more difficult implementations. From there, assess the quantity of necessary Device Guard policies. The easiest solution is to lock down your entire enterprise, but it might not fit your individual departments’ needs.
1.**Group devices into similar functions**. Categorize machines into the groups described in the [Device Guard deployment scenarios](#device-guard-deployment-scenarios) section. This begins the roadmap for your Device Guard deployment and provides groups of easier and more difficult implementations. From there, assess the quantity of necessary Device Guard policies. The easiest solution is to lock down your entire enterprise, but it might not fit your individual departments’ needs.
To discover an appropriate number of policies for your organization, try to separate the defined groups into departments or roles. Then ask some questions: What software does each department or role need to do their job? Should they be able to install and run other departments’ software? Do we need to create a base code integrity policy that aligns with our application catalog? Should users be able to install any application or only choose from an “allowed” list? Do we allow users to use their own peripheral devices? These questions will help you discover the number of necessary policies for your organization. Finally, try to focus on which people or departments would require an additional level of privileges. For example, should department x be able to install and run application xyz, even though no other department does? If the answer is yes and justifiable, you will need a secondary code integrity policy for that group. If not, you will likely be able to merge several policies to simplify management. For more information about configurable code integrity policies, see the [Code integrity policies](#code-integrity-policies) section.
2.**Create code integrity policies from “golden” PCs**. After you create the groups of devices, you can create code integrity policies to align with those groups, similar to the way you would manage corporate images. When you have separated these groups and set up golden PCs that mimic the software and hardware those individual groups require, create code integrity policies from each of them. After you create these, you can merge these code integrity policies to create a master policy, or you can manage and deploy them individually. For step-by-step instructions about how to create code integrity policies, see the [Create code integrity policies from golden PCs](#create-code-golden) section.
2.**Create code integrity policies from “golden” PCs**. After you create the groups of devices, you can create code integrity policies to align with those groups, similar to the way you would manage corporate images. When you have separated these groups and set up golden PCs that mimic the software and hardware those individual groups require, create code integrity policies from each of them. After you create these, you can merge these code integrity policies to create a master policy, or you can manage and deploy them individually. For step-by-step instructions about how to create code integrity policies, see the [Create code integrity policies from golden PCs](#create-code-integrity-policies-from-golden-pcs) section.
3.**Audit and merge code integrity policies**. Microsoft recommends that you test code integrity policies in audit mode before you enforce them. Audit mode allows administrators to run the code integrity policy on a system but not actually block anything. Rather than not allowing applications to run, events are logged with each exception to the policy. This way, you can easily highlight any issues that were not discovered during the initial scan. You can create additional code integrity policies by using the audit events and merge them into the existing policy. For more information about how to audit code integrity policies, see the [Audit code integrity policies](#audit-code-integrity) section.
3.**Audit and merge code integrity policies**. Microsoft recommends that you test code integrity policies in audit mode before you enforce them. Audit mode allows administrators to run the code integrity policy on a system but not actually block anything. Rather than not allowing applications to run, events are logged with each exception to the policy. This way, you can easily highlight any issues that were not discovered during the initial scan. You can create additional code integrity policies by using the audit events and merge them into the existing policy. For more information about how to audit code integrity policies, see the [Audit code integrity policies](#audit-code-integrity-policies) section.
4.**Assess LOB applications that are currently unsigned, and create a catalog file for them**. Catalog files allow organizations to sign applications that do not currently possess digitally signed binaries or applications that a customer would want to add a secondary signature to. These applications can be in-house applications or from third parties, and the process does not require any repackaging of the application. When you create code integrity policies at a rule level above hash values, you will not discover unsigned applications. To include these applications in your code integrity policies, simply create, sign, and deploy a catalog file. For information about catalog files, see the [Catalog files](#catalog-files) section.
5.**Enable desired hardware security features**. Each type of device found in the [Device Guard deployment scenarios](#device-guard-deployment) section takes advantage of different software and hardware integrity configurations. You should assess hardware-based security features separately from code integrity policies because they provide complementary functionality. For information about how to configure Device Guard hardware-based security features, see the [Configure hardware-based security features](#configure-hardware) section.
5.**Enable desired hardware security features**. Each type of device found in the [Device Guard deployment scenarios](#device-guard-deployment-scenarios) section takes advantage of different software and hardware integrity configurations. You should assess hardware-based security features separately from code integrity policies because they provide complementary functionality. For information about how to configure Device Guard hardware-based security features, see the [Configure hardware-based security features](#configure-hardware-based-security-features) section.
6.**Deploy code integrity policies and catalog files**. After you have created and signed the necessary catalog files and created and audited code integrity policies, you are ready to deploy them in phases. Microsoft strongly recommends that you deploy these components to a test group of users, even after your IT organization has tested and vetted them. This provides a final quality control validation before you deploy the catalog files and policies more broadly. For information about how to deploy catalog files with Group Policy, see the [Deploy catalog files with Group Policy](#deploy-cat-gp) section. For additional information about how to deploy code integrity policies, see the [Deploy code integrity policies with Group Policy](#deploy-manage-code-gp) section.
6.**Deploy code integrity policies and catalog files**. After you have created and signed the necessary catalog files and created and audited code integrity policies, you are ready to deploy them in phases. Microsoft strongly recommends that you deploy these components to a test group of users, even after your IT organization has tested and vetted them. This provides a final quality control validation before you deploy the catalog files and policies more broadly. For information about how to deploy catalog files with Group Policy, see the [Deploy catalog files with Group Policy](#deploy-catalog-files-with-group-policy) section. For additional information about how to deploy code integrity policies, see the [Deploy code integrity policies with Group Policy](#deploy-code-integrity-policies-with-group-policy) section.
To help simplify the deployment of Device Guard to your organization, Microsoft recommends that you group devices into the deployment scenarios described in this section. Device Guard is not a feature that organizations will just simply “turn on”; rather, it typically requires a phased implementation approach. To see where these scenarios fit into an overall Device Guard deployment approach, see the [Approach to enterprise code integrity deployment](#approach-enterprise) section.
To help simplify the deployment of Device Guard to your organization, Microsoft recommends that you group devices into the deployment scenarios described in this section. Device Guard is not a feature that organizations will just simply “turn on”; rather, it typically requires a phased implementation approach. To see where these scenarios fit into an overall Device Guard deployment approach, see the [Approach to enterprise code integrity deployment](#approach-to-enterprise-code-integrity-deployment) section.
**Fixed-workload devices**
@ -131,8 +118,6 @@ Device Guard components that are applicable to fixed-workload devices include:
- KMCI VBS protection
<!-- -->
- Enforced UMCI policy
**Fully managed devices**
@ -163,14 +148,11 @@ Device Guard is not a good way to manage devices in a Bring Your Own Device (BYO
## Code signing adoption
Code signing is crucial to the successful implementation of configurable code integrity policies. These policies can trust the signing certificates from both independent software vendors and customers. In Windows10, all Windows Store applications are signed. Also, you can easily trust any other signed application by adding the signing certificate to the code integrity policy.
For unsigned applications, customers have multiple options for signing them so that code integrity policies can trust them. The first option is traditional embedded code signing. Organizations that have in-house development teams can incorporate binary code signing into their application development process, and then simply add the signing certificate to their code integrity policies. The second option for signing unsigned applications is to use catalog files. In Windows10, customers have the ability to create catalog files as they monitor the installation and initial run of an application. For more information about signing existing unsigned LOB applications or third-party applications, see the [Existing line-of-business applications](#existing-lob) section.
For unsigned applications, customers have multiple options for signing them so that code integrity policies can trust them. The first option is traditional embedded code signing. Organizations that have in-house development teams can incorporate binary code signing into their application development process, and then simply add the signing certificate to their code integrity policies. The second option for signing unsigned applications is to use catalog files. In Windows10, customers have the ability to create catalog files as they monitor the installation and initial run of an application. For more information about signing existing unsigned LOB applications or third-party applications, see the [Existing line-of-business applications](#existing-line-of-business-applications) section.
### <a href="" id="existing-lob"></a>
**Existing line-of-business applications**
### Existing line-of-business applications
Until now, existing LOB applications were difficult to trust if they were signed by a source other than the Windows Store or not signed at all. With Windows10, signing your existing LOB and third-party unsigned applications is simplified. This new signing method does not require that applications be repackaged in any way. With catalog files, administrators can sign these unsigned applications simply by monitoring for an installation and initial startup. By using this monitoring information, an administrator can generate a catalog file. Catalog files are simply Secure Hash Algorithm 2 (SHA2) hash lists of discovered binaries. These binaries’ hash values are updated every time an application is updated and therefore require an updated catalog file. For simplified administration, consider incorporating embedded code signing into your application development process. For more information about how to generate catalog files, see the [Catalog files](#catalog-files) section.
@ -178,17 +160,16 @@ Until now, existing LOB applications were difficult to trust if they were signed
Catalog files are lists of individual binaries’ hash values. If the scanned application is updated, you will need to create a new catalog file. That said, binary signing is still highly recommended for any future applications so that no catalog files are needed.
When you create a catalog file, you must sign it by using enterprise public key infrastructure (PKI), or a purchased code signing certificate. When signed, code integrity policies can trust the signer or signing certificate of those files. For information about catalog file signing, see the [Catalog files](#catalog-files) section.
**Application development**
Although in-house applications can be signed after packaging by using catalog files, Microsoft strongly recommends that embedded code signing be incorporated into your application development process. When signing applications, simply add the code signing certificate used to sign your applications to your code integrity policy. This ensures that your code integrity policy will trust any future application that is signed with that certificate. Embedding code signing into any in-house application development process is beneficial to your IT organization as you implement code integrity policies.
Careful consideration about which hardware vendor and specific models to purchase during your next hardware refresh is vitally important to the success of your organization’s Device Guard implementation efforts. In alignment with your current hardware life cycle, consider the process that is discussed in the [Approach enterprise code integrity deployment](#approach-enterprise) section when you determine the appropriate order of hardware replacement in your organization. Device Guard should be deployed in phases; therefore, you have time to methodically plan for its implementation.
Careful consideration about which hardware vendor and specific models to purchase during your next hardware refresh is vitally important to the success of your organization’s Device Guard implementation efforts. In alignment with your current hardware life cycle, consider the process that is discussed in the [Approach enterprise code integrity deployment](#approach-enterprise-code-integrity-deployment) section when you determine the appropriate order of hardware replacement in your organization. Device Guard should be deployed in phases; therefore, you have time to methodically plan for its implementation.
Different hardware features are required to implement the various features of Device Guard. There will likely be some individual features that you will be able to enable with your current hardware and some that you will not. However, for organizations that want to implement Device Guard in its entirety, several advanced hardware features will be required. For additional details about the hardware features that are required for Device Guard components, see the following table.
@ -251,57 +232,47 @@ Different hardware features are required to implement the various features of De
In this section, you learn about the following topics:
- [Configure hardware-based security features](#configure-hardware). This section explains how to enable the hardware-based security features in Device Guard. Also, you verify that the features are enabled by using both Windows Management Infrastructure (WMI) and Msinfo32.exe.
- [Configure hardware-based security features](#configure-hardware-based-security-features). This section explains how to enable the hardware-based security features in Device Guard. Also, you verify that the features are enabled by using both Windows Management Infrastructure (WMI) and Msinfo32.exe.
- [Catalog files](#catalog-files). In this section, you create, sign, and deploy catalog files. You deploy the catalog files by using both Group Policy and System Center Configuration Manager. Also, you use System Center Configuration Manager to inventory the deployed catalog files for reporting purposes.
- [Code integrity policies](#code-integrity-policies). This section provides information on how to create, audit, service, merge, deploy, and remove signed and unsigned configurable code integrity policies.
## <a href="" id="configure-hardware"></a>Configure hardware-based security features
## Configure hardware-based security features
Hardware-based security features make up a large part of Device Guard security offerings. VBS reinforces the most important feature of Device Guard: configurable code integrity. There are three steps to configure hardware-based security features in Device Guard:
1.**Verify that hardware requirements are met and enabled**. Verify that your client machines possess the necessary hardware to run these features. A list of hardware requirements for the hardware-based security features is available in the [Hardware considerations](#hardware) section.
1.**Verify that hardware requirements are met and enabled**. Verify that your client machines possess the necessary hardware to run these features. A list of hardware requirements for the hardware-based security features is available in the [Hardware considerations](#hardware-considerations) section.
2.**Enable the necessary Windows features**. There are several ways to enable the Windows features required for hardware-based security. For details on which Windows features are needed, see the [Windows feature requirements for virtualization-based security](#vb-security) section.
2.**Enable the necessary Windows features**. There are several ways to enable the Windows features required for hardware-based security. For details on which Windows features are needed, see the [Windows feature requirements for virtualization-based security](#windows-feature-requirements-for-virtualization-based-security) section.
3.**Enable desired features**. When the necessary hardware and Windows features have been enabled, you are ready to enable the desired hardware-based security features. For UEFI Secure Boot, see the [Enable UEFI Secure Boot](#enable-secureboot) section. For information about how to enable VBS protection of the KMCI service, see the [Enable virtualization-based protection of kernel mode code integrity](#enable-virtualbased) section. Finally, for information about how to enable Credential Guard, see the [Enable Credential Guard](#enable-cg) section.
3.**Enable desired features**. When the necessary hardware and Windows features have been enabled, you are ready to enable the desired hardware-based security features. For UEFI Secure Boot, see the [Enable UEFI Secure Boot](#enable-unified-extensible-interface-secure-boot) section. For information about how to enable VBS protection of the KMCI service, see the [Enable virtualization-based protection of kernel mode code integrity](#enable-virtualbased) section. Finally, for information about how to enable Credential Guard, see the [Enable Credential Guard](#enable-credential-guard) section.
### <a href="" id="vb-security"></a>
### Windows feature requirements for virtualization-based security
**Windows feature requirements for virtualization-based security**
In addition to the hardware requirements found in the [Hardware considerations](#hardware) section, you must enable certain operating system features before you can enable VBS: Microsoft Hyper-V and isolated user mode (shown in Figure 1).
In addition to the hardware requirements found in the [Hardware considerations](#hardware-considerations) section, you must enable certain operating system features before you can enable VBS: Microsoft Hyper-V and isolated user mode (shown in Figure 1).
**Note**
You can configure these features manually by using Windows PowerShell or Deployment Image Servicing and Management. For specific information about these methods, refer to the [Credential Guard documentation](http://go.microsoft.com/fwlink/p/?LinkId=624529).
Figure 1. Enable operating system features for VBS
After you enable these features, you can configure any hardware-based security features you want. For information about how to enable virtualization-based protection of kernel-mode code integrity, see the [Enable virtualization-based protection of kernel-mode code integrity](#enable-virtualbased) section. For information about how to enable UEFI Secure Boot, see the [Enable Unified Extensible Firmware Interface Secure Boot](#enable-secureboot) section. Finally, for additional information about how to enable Credential Guard, see the [Enable Credential Guard](#enable-cg) section.
After you enable these features, you can configure any hardware-based security features you want. For information about how to enable virtualization-based protection of kernel-mode code integrity, see the [Enable virtualization-based protection of kernel-mode code integrity](#enable-virtualization-based-protection-of-kernel-mode-code-integrity) section. For information about how to enable UEFI Secure Boot, see the [Enable UEFI Secure Boot](#enable-unified-extensible-interface-secure-boot) section. Finally, for additional information about how to enable Credential Guard, see the [Enable Credential Guard](#enable-credential-guard) section.
Before you begin this process, verify that the target device meets the hardware requirements for UEFI Secure Boot that are laid out in the [Hardware considerations](#hardware) section. There are two options to configure UEFI Secure Boot: manual configuration of the appropriate registry keys and Group Policy deployment. Complete the following steps to manually configure UEFI Secure Boot on a computer running Windows 10:
Before you begin this process, verify that the target device meets the hardware requirements for UEFI Secure Boot that are laid out in the [Hardware considerations](#hardware-considerations) section. There are two options to configure UEFI Secure Boot: manual configuration of the appropriate registry keys and Group Policy deployment. Complete the following steps to manually configure UEFI Secure Boot on a computer running Windows 10:
**Note**
There are two platform security levels for Secure Boot: stand-alone Secure Boot and Secure Boot with DMA protection. DMA protection provides additional memory protection but will be enabled only on systems whose processors include DMA protection (IOMMU) technologies. Without the presence of IOMMUs and with DMA protection disabled, customers will lose protection from driver-based attacks.
1. Navigate to the **HKEY\_LOCAL\_MACHINE\\SYSTEM\\CurrentControlSet\\Control\\DeviceGuard** registry subkey.
@ -320,8 +291,6 @@ Unfortunately, it would be time consuming to perform these steps manually on eve
**Note**
Microsoft recommends that you test-enable this feature on a group of test machines before you deploy it to machines that are currently deployed to users.
**Use Group Policy to deploy Secure Boot**
<ahref=""id="bkmk-depsecureboot"></a>
@ -358,17 +327,13 @@ Microsoft recommends that you test-enable this feature on a group of test machin
Processed Device Guard policies are logged in event viewer at Application and Services Logs\\Microsoft\\Windows\\DeviceGuard-GPEXT\\Operational. When the **Turn On Virtualization Based Security** policy is successfully processed, event ID 7000 is logged, which contains the selected settings within the policy.
### <a href="" id="enable-virtualbased"></a>
### Enablevirtualization-based security of kernel-mode code integrity
**Enable virtualization-based security of kernel-mode code integrity**
Before you begin this process, verify that the desired computer meets the hardware requirements for VBS found in the [Hardware considerations](#hardware) section, and enable the Windows features discussed in the [Virtualization-based security Windows feature requirements](#vb-security) section. When validated, you can enable virtualization-based protection of KMCI in one of two ways: manual configuration of the appropriate registry subkeys and Group Policy deployment.
Before you begin this process, verify that the desired computer meets the hardware requirements for VBS found in the [Hardware considerations](#hardware-considerations) section, and enable the Windows features discussed in the [Virtualization-based security Windows feature requirements](#virtualization-based-security-windows-featurerrequirements) section. When validated, you can enable virtualization-based protection of KMCI in one of two ways: manual configuration of the appropriate registry subkeys and Group Policy deployment.
**Note**
All drivers on the system must be compatible with virtualization-based protection of code integrity; otherwise, your system may fail. Microsoft recommends that you enable this feature on a group of test machines before you enable it on deployed machines.
To configure virtualization-based protection of KMCI manually:
1. Navigate to the **HKEY\_LOCAL\_MACHINE\\SYSTEM\\CurrentControlSet\\Control\\DeviceGuard** registry subkey.
@ -382,8 +347,6 @@ It would be time consuming to perform these steps manually on every protected ma
**Note**
Microsoft recommends that you test-enable this feature on a group of test computers before you deploy it to machines that are currently deployed to users. If untested, there is a possibility that this feature can cause system instability and ultimately cause the client operating system to fail.
To use Group Policy to configure VBS of KMCI:
1. Create a new GPO: Right-click the OU to which you want to link the GPO, and then click **Create a GPO in this domain, and Link it here**.
@ -416,13 +379,11 @@ To use Group Policy to configure VBS of KMCI:
Processed Device Guard policies are logged in event viewer under Application and Services Logs\\Microsoft\\Windows\\DeviceGuard-GPEXT\\Operational. When the **Turn On Virtualization Based Security** policy has been successfully processed, event ID 7000 is logged, which contains the selected settings within the policy.
### <a href="" id="enable-cg"></a>
**Enable Credential Guard**
### Enable Credential Guard
Credential Guard provides an additional layer of credential protection specifically for domain users by storing the credentials within the virtualized container, away from both the kernel and user mode operating system. This makes it difficult for even a compromised system to obtain access to the credentials. In addition to the client-side enablement of Credential Guard, you can deploy additional mitigations at both the Certification Authority and domain controller level to prevent credential theft. Microsoft will be releasing details about these additional mitigations in the future.
Before you begin this process, verify that the desired system meets the hardware requirements for VBS found in the [Hardware considerations](#hardware) section, and that you have enabled the Windows features laid out in the [Virtualization-based security Windows feature requirements](#vb-security) section. When validated, you can enable Credential Guard manually, by configuring the appropriate registry subkeys, or through Group Policy deployment.
Before you begin this process, verify that the desired system meets the hardware requirements for VBS found in the [Hardware considerations](#hardware) section, and that you have enabled the Windows features laid out in the [Virtualization-based security Windows feature requirements](#virtualization-based-security-windows-feature-requirements) section. When validated, you can enable Credential Guard manually, by configuring the appropriate registry subkeys, or through Group Policy deployment.
To configure VBS of Credential Guard manually:
@ -437,8 +398,6 @@ To avoid spending an unnecessary amount of time in manual deployments, use Group
**Note**
Microsoft recommends that you enable Credential Guard before you join a machine to the domain to ensure that all credentials are properly protected. Setting the appropriate registry subkeys during your imaging process would be ideal to achieve this protection.
To use Group Policy to enable Credential Guard:
1. Create a new GPO: right-click the OU to which you want to link the GPO, and then click **Create a GPO in this domain, and Link it here** .
@ -470,8 +429,6 @@ To use Group Policy to enable Credential Guard:
**Note**
The default platform security level is **Secure Boot**. If IOMMUs are available within the protected machines, it is recommended that you select **Secure Boot and DMA Protection** to maximize the mitigations that are available through Credential Guard.
7. Check the test client event log for Device Guard GPOs.
Another method to determine the available and enabled Device Guard features is to run msinfo32.exe from an elevated PowerShell session. When you run this program, the Device Guard properties are displayed at the bottom of the **System Summary** section, as shown in Figure 11.
@ -585,43 +540,34 @@ Figure 11. Device Guard properties in the System Summary
## Catalog files
Enforcement of Device Guard on a system requires that every trusted application have a signature or its binary hashes added to the code integrity policy. For many organizations, this can be an issue when considering unsigned LOB applications. To avoid the requirement that organizations repackage and sign these applications, Windows10 includes a tool called Package Inspector that monitors an installation process for any deployed and executed binary files. If the tool discovers such files, it itemizes them in a catalog file. These catalog files offer you a way to trust your existing unsigned applications, whether developed in house or by a third party, as well as trust signed applications for which you do not want to trust the signer but rather the specific application. When created, these files can be signed, the signing certificates added to your existing code integrity policies, and the catalog files themselves distributed to the clients.
**Note**
The Enterprise edition of Windows10 or Windows Server 2016 is required to create and use catalog files.
### <a href="" id="create-catalog-files"></a>
**Create catalog files**
### Create catalog files
The creation of catalog files is the first step to add an unsigned application to a code integrity policy. To create a catalog file, copy each of the following commands into an elevated Windows PowerShell session, and then complete the steps:
**Note**
When you establish a naming convention it makes it easier to detect deployed catalog files in the future. In this guide, you will use *\*-Contoso.cat* as the naming convention. For more information about why this practice is helpful to inventory or detect catalog files, see the [Inventory catalog files with System Center Configuration Manager](#inventory-cat-sccm) section.
When you establish a naming convention it makes it easier to detect deployed catalog files in the future. In this guide, you will use *\*-Contoso.cat* as the naming convention. For more information about why this practice is helpful to inventory or detect catalog files, see the [Inventory catalog files with System Center Configuration Manager](#inventory-catalog-files-with-system-center-configuration-manager) section.
1. Be sure that a code integrity policy is currently running in audit mode.
Package Inspector does not always detect installation files that have been removed from the machine during the installation process. To ensure that these binaries are also trusted, the code integrity policy that you created and audited in the [Create code integrity policies from golden PCs](#create-code-golden) and [Audit code integrity policies](#audit-code-integrity) sections should be deployed, in audit mode, to the system on which you are running Package Inspector.
Package Inspector does not always detect installation files that have been removed from the machine during the installation process. To ensure that these binaries are also trusted, the code integrity policy that you created and audited in the [Create code integrity policies from golden PCs](#create-code-integrity-policies-from-golden-pcs) and [Audit code integrity policies](#audit-code-integrity-policies) sections should be deployed, in audit mode, to the system on which you are running Package Inspector.
**Note**
This process should **not** be performed on a system running an enforced Device Guard policy, only with a policy running in audit mode. If a policy is currently being enforced, you will not be able to install and run the application.
2. Start Package Inspector, and then scan drive C:
`PackageInspector.exe Start C:`
**Note**
Package inspector can monitor installations on any local drive. In this example, we install the application on drive C, but any other drive can be used.
3. Copy the installation media to drive C.
By copying the installation media to drive C, you ensure that Package Inspector detects and catalogs the actual installer. If you skip this step, the future code integrity policy may trust the application to run but not be installed.
@ -648,13 +594,9 @@ When you establish a naming convention it makes it easier to detect deployed cat
**Note**
This scan catalogs the hash values for each discovered binary file. If the applications that were scanned are updated, complete this process again to trust the new binaries’ hash values.
When finished, the files will be saved to your desktop. To trust this catalog file within a code integrity policy, the catalog must first be signed. Then, the signing certificate can be included in the code integrity policy, and the catalog file can be distributed to the individual client machines. Catalog files can be signed by using a certificate and SignTool.exe, a free tool available in the Windows SDK. For more information about signing catalog files with SignTool.exe, see the [Catalog signing with SignTool.exe](#catalog-signing-with-signtool.exe) section.
When finished, the files will be saved to your desktop. To trust this catalog file within a code integrity policy, the catalog must first be signed. Then, the signing certificate can be included in the code integrity policy, and the catalog file can be distributed to the individual client machines. Catalog files can be signed by using a certificate and SignTool.exe, a free tool available in the Windows SDK. For more information about signing catalog files with SignTool.exe, see the [Catalog signing with SignTool.exe](#catsign-signtool) section.
### <a href="" id="catsign-signtool"></a>
**Catalog signing with SignTool.exe**
### Catalog signing with SignTool.exe
Device Guard makes it easy for organizations to sign and trust existing unsigned LOB applications. In this section, you sign a catalog file you generated in a previous section by using PackageInspector.exe. For information about how to create catalog files, see the [Create catalog files](#create-catalog-files) section. In this example, you need the following:
@ -664,62 +606,28 @@ Device Guard makes it easy for organizations to sign and trust existing unsigned
If you do not have a code signing certificate, please see the [Create a Device Guard code signing certificate](#create-dg-code) section for a walkthrough of how to create one. In addition to using the certificate you create in the Create a Device Guard code signing certificate section, this example signs the catalog file that you created in the [Create catalog files](#create-catalog-files) section. If you are using an alternate certificate or catalog file, update the following steps with the appropriate variables and certificate. To sign the existing catalog file, copy each of the following commands into an elevated Windows PowerShell session:
If you do not have a code signing certificate, please see the [Create a Device Guard code signing certificate](#create-a-device-guard-code-signing-certificate) section for a walkthrough of how to create one. In addition to using the certificate you create in the Create a Device Guard code signing certificate section, this example signs the catalog file that you created in the [Create catalog files](#create-catalog-files) section. If you are using an alternate certificate or catalog file, update the following steps with the appropriate variables and certificate. To sign the existing catalog file, copy each of the following commands into an elevated Windows PowerShell session:
In this example, you use the catalog file you created in the [Create catalog files](#create-catalog-files) section. If you are signing another catalog file, be sure to update the *$ExamplePath* and *$CatFileName* variables with the correct information.
**Note**
In this example, you use the catalog file you created in the [Create catalog files](#create-catalog-files) section. If you are signing another catalog file, be sure to update the *$ExamplePath* and *$CatFileName* variables with the correct information.
2. Import the code signing certificate. Import the code signing certificate that will be used to sign the catalog file to the signing user’s personal store. In this example, you use the certificate that you created in the [Create a Device Guard code signing certificate](#create-dg-code) section.
2. Import the code signing certificate. Import the code signing certificate that will be used to sign the catalog file to the signing user’s personal store. In this example, you use the certificate that you created in the [Create a Device Guard code signing certificate](#create-a-device-guard-code-signing-certificate) section.
3. Sign the catalog file with Signtool.exe:
<spancodelanguage=""></span>
<table>
<colgroup>
<colwidth="100%"/>
</colgroup>
<tbody>
<trclass="odd">
<tdalign="left"><pre><code><Path to signtool.exe> sign /n "ContosoDGSigningCert" /fd sha256 /v $CatFileName</code></pre></td>
</tr>
</tbody>
</table>
`<path to signtool.exe> sign /n "ContosoDGSigningCert" /fd sha256 /v $CatFileName`
**Note**
The *<Path to signtool.exe>* variable should be the full path to the Signtool.exe utility. *ContosoDGSigningCert* is the subject name of the certificate that you will use to sign the catalog file. This certificate should be imported to your personal certificate store on the machine on which you are attempting to sign the catalog file.
**Note**
The *<Path to signtool.exe>* variable should be the full path to the Signtool.exe utility. *ContosoDGSigningCert* is the subject name of the certificate that you will use to sign the catalog file. This certificate should be imported to your personal certificate store on the machine on which you are attempting to sign the catalog file.
**Note**
For additional information about Signtool.exe and all additional switches, visit [MSDN Sign Tool page](http://go.microsoft.com/fwlink/p/?LinkId=624163).
**Note**
For additional information about Signtool.exe and all additional switches, visit [MSDN Sign Tool page](http://go.microsoft.com/fwlink/p/?LinkId=624163).
@ -733,31 +641,25 @@ If you do not have a code signing certificate, please see the [Create a Device G
For testing purposes, you can manually copy signed catalog files to their intended folder. For large-scale implementations, Microsoft recommends that you use Group Policy File Preferences to copy the appropriate catalog files to all desired machines or an enterprise systems management product such as System Center Configuration Manager. Doing this simplifies the management of catalog versions, as well.
### <a href="" id="deploy-cat-gp"></a>
**Deploy catalog files with Group Policy**
### Deploycatalog files with Group Policy
To simplify the management of catalog files, you can use Group Policy preferences to deploy catalog files to the appropriate PCs in your organization. The following process walks you through the deployment of a signed catalog file called LOBApp-Contoso.cat to a test OU called DG Enabled PCs with a GPO called **Contoso DG Catalog File GPO Test**.
**Note**
This walkthrough requires that you have previously created a signed catalog file and have a Windows 10 client PC on which to test a Group Policy deployment. For more information about how to create and sign a catalog file, see the [Catalog files](#catalog-files) section.
To deploy a catalog file with Group Policy:
1. From either a domain controller or a client PC that has Remote Server Administration Tools (RSAT) installed, open the Group Policy Management Console (GPMC) by running **GPMC.MSC** or by searching for Group Policy Management.
2. Create a new GPO: right-click the DG Enabled PCs OU, and then click **Create a GPO in this domain, and Link it here**, as shown in Figure 13.
**Note**
The DG Enabled PCs OU is just an example of where to link the test GPO that you created in this section. You can use any OU name. Also, security group filtering is an option when you consider policy partitioning options based on the strategy discussed in the [Approach enterprise code integrity deployment](#approach-enterprise) section.
**Note**
The DG Enabled PCs OU is just an example of where to link the test GPO that you created in this section. You can use any OU name. Also, security group filtering is an option when you consider policy partitioning options based on the strategy discussed in the [Approach enterprise code integrity deployment](#approach-enterprise-code-integrity-deployment) section.
Figure 13. Create a new GPO
Figure 13. Create a new GPO
3. Name the new GPO **Contoso DG Catalog File GPO Test**.
@ -796,17 +698,13 @@ To deploy a catalog file with Group Policy:
12. Close the Group Policy Management Editor, and then update the policy on the test Windows 10 machine by running GPUpdate.exe. When the policy has been updated, verify that the catalog file exists in C:\\Windows\\System32\\catroot\\{F750E6C3-38EE-11D1-85E5-00C04FC295EE} on the Windows 10 machine.
### <a href="" id="deploy-cat-sccm"></a>
**Deploy catalog files with System Center Configuration Manager**
### Deploycatalog files with System Center Configuration Manager
As an alternative to Group Policy, you can use System Center Configuration Manager to deploy catalog files to the managed machines in your environment. This approach can simplify the deployment and management of multiple catalog files as well as provide reporting around which catalog each client or collection has deployed. In addition to the deployment of these files, System Center Configuration Manager can also be used to inventory the currently deployed catalog files for reporting and compliance purposes. Complete the following steps to create a new deployment package for catalog files:
**Note**
The following example uses a network share named \\\\Shares\\CatalogShare as a source for the catalog files. If you have collection specific catalog files, or prefer to deploy them individually, use whichever folder structure works best for your organization.
1. Open the Configuration Manager console, and select the Software Library workspace.
2. Navigate to Overview\\Application Management, right-click **Packages**, and then click **Create Package**.
@ -873,17 +771,13 @@ After you create the deployment package, deploy it to a collection so that the c
11. Close the wizard.
### <a href="" id="inventory-cat-sccm"></a>
**Inventory catalog files with System Center Configuration Manager**
### Inventorycatalog files with System Center Configuration Manager
When catalog files have been deployed to the machines within your environment, whether by using Group Policy or System Center Configuration Manager, you can inventory them with the software inventory feature of System Center Configuration Manager. The following process walks you through the enablement of software inventory to discover catalog files on your managed systems through the creation and deployment of a new client settings policy.
**Note**
A standard naming convention for your catalog files will significantly simplify the catalog file software inventory process. In this example, *-Contoso* has been added to all catalog file names.
1. Open the Configuration Manager console, and select the Administration workspace.
2. Navigate to **Overview\\Client Settings**, right-click **Client Settings**, and then click **Create Custom Client Device Settings**.
@ -937,25 +831,19 @@ If nothing is displayed in this view, navigate to Software\\Last Software Scan i
## Code integrity policies
Code integrity policies maintain the standards by which a computer running Windows10 determines whether an application is trustworthy and can be run. For an overview of code integrity, see the [Configurable code integrity](#config-code) section.
Code integrity policies maintain the standards by which a computer running Windows10 determines whether an application is trustworthy and can be run. For an overview of code integrity, see the [Configurable code integrity](#configurable-code-integrity) section.
A common system imaging practice in today’s IT organization is to establish a “golden” image as a reference for what an ideal system should look like, and then use that image to clone additional company assets. Code integrity policies follow a similar methodology, that begins with the establishment of a golden PC. Like when imaging, you can have multiple golden PCs based on model, department, application set, and so on. Although the thought process around the creation of code integrity policies is similar to imaging, these policies should be maintained independently. Assess the necessity of additional code integrity policies based on what should be allowed to be installed and run and for whom.
**Note**
Each machine can have only **one** code integrity policy at a time. Whichever way you deploy this policy, it is renamed to SIPolicy.p7b and copied to C:\\Windows\\System32\\CodeIntegrity. Keep this in mind when you create your code integrity policies.
Optionally, code integrity policies can align with your software catalog as well as any IT department–approved applications. One simple method to implement code integrity policies is to use existing images to create one master code integrity policy. You do so by creating a code integrity policy from each image, and then by merging the policies. This way, what is installed on all of those images will be allowed to run, should the applications be installed on a computer based on a different image. Alternatively, you may choose to create a base applications policy and add policies based on the computer’s role or department. Organizations have a choice of how their policies are created, merged or serviced, and managed.
**Note**
The following section assumes that you will deploy code integrity policies as part of your Device Guard deployment. Alternatively, configurable code integrity is available without the enablement of Device Guard.
Code integrity policies consist of several components. The two major components, which are configurable, are called *policy rules* and *file rules*, respectively. Code integrity policy rules are options that the code integrity policy creator can specify on the policy. These options include the enablement of audit mode, UMCI, and so on. You can modify these options in a new or existing code integrity policy. File rules are the level to which the code integrity policy scan ties each binary trust. For example, the hash level is going to itemize each discovered hash on the system within the generated code integrity policy. This way, when a binary prepares to run, the code integrity service will validate its hash value against the trusted hashes found in the code integrity policy. Based on that result, the binary will or will not be allowed to run.
@ -973,58 +861,48 @@ You can set several rule options within a code integrity policy. Table 2 lists e
| **0 Enabled:UMCI** | Code integrity policies restrict both kernel-mode and user-mode binaries. By default, only kernel-mode binaries are restricted. Enabling this rule option validates user mode executables and scripts. |
| **1 Enabled:Boot Menu Protection** | This option is not currently supported. |
| **2 Required:WHQL** | By default, legacy drivers that are not Windows Hardware Quality Labs (WHQL) signed are allowed to execute. Enabling this rule requires that every executed driver is WHQL signed and removes legacy driver support. Going forward, every new Windows 10–compatible driver must be WHQL certified. |
| **3 Enabled:Audit Mode (Default)** | Enables the execution of binaries outside of the code integrity policy but logs each occurrence in the CodeIntegrity event log, which can be used to update the existing policy before enforcement. To enforce a code integrity policy, remove this option. |
| **4 Disabled:Flight Signing** | If enabled, code integrity policies will not trust flightroot-signed binaries. This would be used in the scenario in which organizations only want to run released binaries, not flighted builds. |
| **5 Enabled:Inherent Default Policy** | This option is not currently supported. |
| **6 Enabled:Unsigned System Integrity Policy (Default)** | Allows the policy to remain unsigned. When this option is removed, the policy must be signed and have UpdatePolicySigners added to the policy to enable future policy modifications. |
| **7 Allowed:Debug Policy Augmented** | This option is not currently supported. |
| **8 Required:EV Signers** | In addition to being WHQL signed, this rule requires that drivers must have been submitted by a partner that has an Extended Verification (EV) certificate. All future Windows 10 and later drivers will meet this requirement. |
| **9 Enabled:Advanced Boot Options Menu** | The F8 preboot menu is disabled by default for all code integrity policies. Setting this rule option allows the F8 menu to appear to physically present users. |
| **10 Enabled:Boot Audit on Failure** | Used when the code integrity policy is in enforcement mode. When a driver fails during startup, the code integrity policy will be placed in audit mode so that Windows will load. Administrators can validate the reason for the failure in the CodeIntegrity event log. |
| Rule option | Description |
|------------ | ----------- |
| **0 Enabled:UMCI** | Code integrity policies restrict both kernel-mode and user-mode binaries. By default, only kernel-mode binaries are restricted. Enabling this rule option validates user mode executables and scripts. |
| **1 Enabled:Boot Menu Protection** | This option is not currently supported. |
| **2 Required:WHQL** | By default, legacy drivers that are not Windows Hardware Quality Labs (WHQL) signed are allowed to execute. Enabling this rule requires that every executed driver is WHQL signed and removes legacy driver support. Going forward, every new Windows 10–compatible driver must be WHQL certified. |
| **3 Enabled:Audit Mode (Default)** | Enables the execution of binaries outside of the code integrity policy but logs each occurrence in the CodeIntegrity event log, which can be used to update the existing policy before enforcement. To enforce a code integrity policy, remove this option. |
| **4 Disabled:Flight Signing** | If enabled, code integrity policies will not trust flightroot-signed binaries. This would be used in the scenario in which organizations only want to run released binaries, not flighted builds. |
| **5 Enabled:Inherent Default Policy** | This option is not currently supported. |
| **6 Enabled:Unsigned System Integrity Policy (Default)** | Allows the policy to remain unsigned. When this option is removed, the policy must be signed and have UpdatePolicySigners added to the policy to enable future policy modifications. |
| **7 Allowed:Debug Policy Augmented** | This option is not currently supported. |
| **8 Required:EV Signers** | In addition to being WHQL signed, this rule requires that drivers must have been submitted by a partner that has an Extended Verification (EV) certificate. All future Windows 10 and later drivers will meet this requirement. |
| **9 Enabled:Advanced Boot Options Menu** | The F8 preboot menu is disabled by default for all code integrity policies. Setting this rule option allows the F8 menu to appear to physically present users. |
| **10 Enabled:Boot Audit on Failure** | Used when the code integrity policy is in enforcement mode. When a driver fails during startup, the code integrity policy will be placed in audit mode so that Windows will load. Administrators can validate the reason for the failure in the CodeIntegrity event log. |
File rule levels allow administrators to specify the level at which they want to trust their applications. This level of trust could be as low as the hash of each binary and as high as a PCA certificate. File rule levels are specified both when you create a new code integrity policy from a scan and when you create a policy from audit events. In addition, to combine rule levels found in multiple policies, you can merge the policies. When merged, code integrity policies combine their file rules. Each file rule level has its benefit and disadvantage. Use Table 3 to select the appropriate protection level for your available administrative resources and Device Guard deployment scenario.
| **Hash** | Specifies individual hash values for each discovered binary. Although this level is specific, it can cause additional administrative overhead to maintain the current product versions’ hash values. Each time a binary is updated, the hash value changes, therefore requiring a policy update. |
| **FileName** | Specifies individual binary file names. Although the hash values for an application are modified when updated, the file names are typically not. This offers less specific security than the hash level but does not typically require a policy update when any binary is modified. |
| **SignedVersion** | This combines the publisher rule with a file version number. This option allows anything from the specified publisher, with a file version at or above the specified version number, to run. |
| **Publisher** | This is a combination of the PCA certificate and the common name (CN) on the leaf certificate. In the scenario that a PCA certificate is used to sign multiple companies’ applications (such as VeriSign), this rule level allows organizations to trust the PCA certificate but only for the company whose name is on the leaf certificate (for example, Intel for device drivers). This level trusts a certificate with a long validity period but only when combined with a trusted leaf certificate. |
| **FilePublisher** | This is a combination of the publisher file rule level and the SignedVersion rule level. Any signed file from the trusted publisher that is the specified version or newer is trusted. |
| **LeafCertificate** | Adds trusted signers at the individual signing certificate level. The benefit of using this level versus the individual hash level is that new versions of the product will have different hash values but typically the same signing certificate. Using this level, no policy update would be needed to run the new version of the application. However, leaf certificates have much shorter validity periods than PCA certificates, so additional administrative overhead is associated with updating the code integrity policy when these certificates expire. |
| **PcaCertificate** | Adds the highest certificate in the provided certificate chain to signers. This is typically one certificate below the root certificate, because the scan does not validate anything above the presented signature by going online or checking local root stores. |
| **RootCertificate** | Currently unsupported. |
| **WHQL** | Trusts binaries if they have been validated and signed by WHQL. This is primarily for kernel binaries. |
| **WHQLPublisher** | This is a combination of the WHQL and the CN on the leaf certificate and is primarily for kernel binaries. |
| **WHQLFilePublisher** | Specifies that the binaries are validated and signed by WHQL, with a specific publisher (WHQLPublisher), and that the binary is the specified version or newer. This is primarily for kernel binaries. |
| Rule level | Description |
|----------- | ----------- |
| **Hash** | Specifies individual hash values for each discovered binary. Although this level is specific, it can cause additional administrative overhead to maintain the current product versions’ hash values. Each time a binary is updated, the hash value changes, therefore requiring a policy update. |
| **FileName** | Specifies individual binary file names. Although the hash values for an application are modified when updated, the file names are typically not. This offers less specific security than the hash level but does not typically require a policy update when any binary is modified. |
| **SignedVersion** | This combines the publisher rule with a file version number. This option allows anything from the specified publisher, with a file version at or above the specified version number, to run. |
| **Publisher** | This is a combination of the PCA certificate and the common name (CN) on the leaf certificate. In the scenario that a PCA certificate is used to sign multiple companies’ applications (such as VeriSign), this rule level allows organizations to trust the PCA certificate but only for the company whose name is on the leaf certificate (for example, Intel for device drivers). This level trusts a certificate with a long validity period but only when combined with a trusted leaf certificate. |
| **FilePublisher** | This is a combination of the publisher file rule level and the SignedVersion rule level. Any signed file from the trusted publisher that is the specified version or newer is trusted. |
| **LeafCertificate** | Adds trusted signers at the individual signing certificate level. The benefit of using this level versus the individual hash level is that new versions of the product will have different hash values but typically the same signing certificate. Using this level, no policy update would be needed to run the new version of the application. However, leaf certificates have much shorter validity periods than PCA certificates, so additional administrative overhead is associated with updating the code integrity policy when these certificates expire. |
| **PcaCertificate** | Adds the highest certificate in the provided certificate chain to signers. This is typically one certificate below the root certificate, because the scan does not validate anything above the presented signature by going online or checking local root stores. |
| **RootCertificate** | Currently unsupported. |
| **WHQL** | Trusts binaries if they have been validated and signed by WHQL. This is primarily for kernel binaries. |
| **WHQLPublisher** | This is a combination of the WHQL and the CN on the leaf certificate and is primarily for kernel binaries. |
| **WHQLFilePublisher** | Specifies that the binaries are validated and signed by WHQL, with a specific publisher (WHQLPublisher), and that the binary is the specified version or newer. This is primarily for kernel binaries. |
**Note**
When you create code integrity policies with the **New-CIPolicy** cmdlet, you can specify a primary file rule level by including the **–Level** parameter. For discovered binaries that cannot be trusted based on the primary file rule criteria, use the **–Fallback** parameter. For example, if the primary file rule level is PCACertificate but you would like to trust the unsigned applications as well, using the Hash rule level as a fallback adds the hash values of binaries that did not have a signing certificate.
### <a href="" id="create-code-golden"></a>
**Create code integrity policies from golden PCs**
### Create code integrity policies from golden PCs
The process to create a golden code integrity policy from a reference system is straightforward. This section outlines the process that is required to successfully create a code integrity policy with Windows PowerShell. First, for this example, you must initiate variables to be used during the creation process. Rather than using variables, you can simply use the full file paths in the command. Next, you create the code integrity policy by scanning the system for installed applications. When created, the policy file is converted to binary format so that Windows can consume its contents.
**Note**
Before you begin this procedure, ensure that the reference PC is clean of viruses or malware. Each piece of installed software should be validated as trustworthy before you create this policy. Also, be sure that any software that you would like to be scanned is installed on the system before you create the code integrity policy.
To create a code integrity policy, copy each of the following commands into an elevated Windows PowerShell session, in order:
1. Initialize variables that you will use:
@ -1039,22 +917,16 @@ To create a code integrity policy, copy each of the following commands into an e
By specifying the *–UserPEs* parameter, rule option **0 Enabled:UMCI** is automatically added to the code integrity policy. If you do not specify this parameter, use the following command to enable UMCI:
**Note**
By specifying the *–UserPEs* parameter, rule option **0 Enabled:UMCI** is automatically added to the code integrity policy. If you do not specify this parameter, use the following command to enable UMCI:
You can add the *–Fallback* parameter to catch any applications not discovered using the primary file rule level specified by the *–Level* parameter. For more information about file rule level options, see the [Code integrity policy rules](#code-integrity-policy-rules) section.
**Note**
You can add the *–Fallback* parameter to catch any applications not discovered using the primary file rule level specified by the *–Level* parameter. For more information about file rule level options, see the [Code integrity policy rules](#code-integrity-policy-rules) section.
**Note**
If you would like to specify the code integrity policy scan to look only at a specific drive, you can do so by using the *–ScanPath* parameter. Without this parameter, as shown in the example, the entire system is scanned.
**Note**
If you would like to specify the code integrity policy scan to look only at a specific drive, you can do so by using the *–ScanPath* parameter. Without this parameter, as shown in the example, the entire system is scanned.
3. Convert the code integrity policy to a binary format:
@ -1063,57 +935,45 @@ To create a code integrity policy, copy each of the following commands into an e
After you complete these steps, the Device Guard binary file (DeviceGuardPolicy.bin) and original .xml file (IntialScan.xml) will be available on your desktop. You can use the binary version as a code integrity policy or sign it for additional security.
**Note**
Microsoft recommends that you keep the original .xml file of the policy for use when you need to merge the code integrity policy with another policy or update its rule options. Alternatively, you would have to create a new policy from a new scan for servicing. For more information about how to merge code integrity policies, see the [Merge code integrity policies](#merge-code-integrity) section.
Microsoft recommends that you keep the original .xml file of the policy for use when you need to merge the code integrity policy with another policy or update its rule options. Alternatively, you would have to create a new policy from a new scan for servicing. For more information about how to merge code integrity policies, see the [Merge code integrity policies](#merge-code-integrity-policies) section.
Microsoft recommends that every code integrity policy be run in audit mode before being enforced. Doing so allows administrators to discover any issues with the policy without receiving error message dialog boxes. For information about how to audit a code integrity policy, see the [Audit code integrity policies](#audit-code-integrity-policies) section.
Microsoft recommends that every code integrity policy be run in audit mode before being enforced. Doing so allows administrators to discover any issues with the policy without receiving error message dialog boxes. For information about how to audit a code integrity policy, see the [Audit code integrity policies](#audit-code-integrity) section.
### <a href="" id="audit-code-integrity"></a>
**Audit code integrity policies**
### Audit code integrity policies
When code integrity policies are run in audit mode, it allows administrators to discover any applications that were missed during an initial policy scan and to identify any new applications that have been installed and run since the original policy was created. While a code integrity policy is running in audit mode, any binary that runs and would have been denied had the policy been enforced is logged in the Applications and Services Logs\\Microsoft\\CodeIntegrity\\Operational event log. When these logged binaries have been validated, they can easily be added to a new code integrity policy. When the new exception policy is created, you can merge it with your existing code integrity policies.
**Note**
Before you begin this process, you need to create a code integrity policy binary file. If you have not already done so, see the [Create a code integrity policy](#create-code-golden) section for a step-by-step walkthrough of the process to create a code integrity policy and convert it to binary format.
Before you begin this process, you need to create a code integrity policy binary file. If you have not already done so, see the [Create a code integrity policy](#create-a-code-integrity-policy) section for a step-by-step walkthrough of the process to create a code integrity policy and convert it to binary format.
To audit a code integrity policy with local policy:
1. Copy the DeviceGuardPolicy.bin file that you created in the [Create code integrity policies from golden PCs](#create-code-golden) section to C:\\Windows\\System32\\CodeIntegrity.
1. Copy the DeviceGuardPolicy.bin file that you created in the [Create code integrity policies from golden PCs](#create-code-integrity-policies-from-golden-pcs) section to C:\\Windows\\System32\\CodeIntegrity.
2. On the system you want to run in audit mode, open the Local Group Policy Editor by running **GPEdit.msc**.
3. Navigate to Computer Configuration\\Administrative Templates\\System\\Device Guard, and then select **Deploy Code Integrity Policy**. Enable this setting by using the file path C:\\Windows\\System32\\CodeIntegrity\\DeviceGuardPolicy.bin, as shown in Figure 22.
**Note**
*DeviceGuardPolicy.bin* is not a required policy name. This name was simply used in the [Create code integrity policies from golden PCs](#create-code-golden) section and so was used here. Also, this policy file does not need to be copied to every system. Alternatively, you can copy the code integrity policies to a file share to which all computer accounts have access.
**Note**
*DeviceGuardPolicy.bin* is not a required policy name. This name was simply used in the [Create code integrity policies from golden PCs](#create-code-golden) section and so was used here. Also, this policy file does not need to be copied to every system. Alternatively, you can copy the code integrity policies to a file share to which all computer accounts have access.
**Note**
Any policy you select here is converted to SIPolicy.p7b when it is deployed to the individual computers.
**Note**
Any policy you select here is converted to SIPolicy.p7b when it is deployed to the individual computers.
Figure 22. Deploy your code integrity policy
Figure 22. Deploy your code integrity policy
**Note**
You may have noticed that the GPO setting references a .p7b file and this policy uses a .bin file. Regardless of the type of policy you deploy (.bin, .p7b, or .p7), they are all converted to SIPolicy.p7b when dropped on the Windows 10 computers. Microsoft recommends that you make your code integrity policies friendly and allow the system to convert the policy names for you. By doing this, it ensures that the policies are easily distinguishable when viewed in a share or any other central repository.
**Note**
You may have noticed that the GPO setting references a .p7b file and this policy uses a .bin file. Regardless of the type of policy you deploy (.bin, .p7b, or .p7), they are all converted to SIPolicy.p7b when dropped on the Windows 10 computers. Microsoft recommends that you make your code integrity policies friendly and allow the system to convert the policy names for you. By doing this, it ensures that the policies are easily distinguishable when viewed in a share or any other central repository.
4. Restart reference system for the code integrity policy to take effect.
5. Monitor the CodeIntegrity event log. While in audit mode, any exception to the deployed code integrity policy will be logged in the Applications and Services Logs\\Microsoft\\CodeIntegrity\\Operational event log, as shown in Figure 23.
Figure 23. Exceptions to the deployed code integrity policy
Figure 23. Exceptions to the deployed code integrity policy
6. Validate any code integrity policy exceptions.
@ -1126,11 +986,7 @@ To audit a code integrity policy with local policy:
**Note**
An alternative method to test a policy is to rename the test file to SIPolicy.p7b and drop it into C:\\Windows\\System32\\CodeIntegrity, rather than deploy it with the local machine policy.
### <a href="" id="create-audit-code"></a>
**Create an audit code integrity policy**
### Create an audit code integrity policy
When you run code integrity policies in audit mode, validate any exceptions and determine whether you will need to add them to the code integrity policy you want to audit. Use the system as you normally would to ensure that any use exceptions are logged. When you are ready to create a code integrity policy from the auditing events, complete the following steps in an elevated Windows PowerShell session:
@ -1142,7 +998,7 @@ When you run code integrity policies in audit mode, validate any exceptions and
2. Analyze audit results.
Before you create a code integrity policy from audit events, Microsoft recommends that each exception be analyzed, as discussed in steps 5 and 6 of the [Audit code integrity policies](#audit-code-integrity) section.
Before you create a code integrity policy from audit events, Microsoft recommends that each exception be analyzed, as discussed in steps 5 and 6 of the [Audit code integrity policies](#audit-code-integrity-policies) section.
3. Generate a new code integrity policy from logged audit events:
@ -1151,25 +1007,17 @@ When you run code integrity policies in audit mode, validate any exceptions and
**Note**
When you create policies from audit events, you should carefully consider the file rule level that you select to trust. In this example, you use the Hash rule level, which should be used as a last resort.
After you complete these steps, the Device Guard audit policy .xml file (DeviceGuardAuditPolicy.xml) will be available on your desktop. You can now use this file to update the existing code integrity policy that you ran in audit mode by merging the two policies. For instructions on how to merge this audit policy with the existing code integrity policy, see the [Merge code integrity policies](#merge-code-integrity) section.
After you complete these steps, the Device Guard audit policy .xml file (DeviceGuardAuditPolicy.xml) will be available on your desktop. You can now use this file to update the existing code integrity policy that you ran in audit mode by merging the two policies. For instructions on how to merge this audit policy with the existing code integrity policy, see the [Merge code integrity policies](#merge-code-integrity-policies) section.
**Note**
You may have noticed that you did not generate a binary version of this policy as you did in the [Create code integrity policies from golden PCs](#create-code-golden) section. This is because code integrity policies created from an audit log are not intended to run as stand-alone policies but rather to update existing code integrity policies.
You may have noticed that you did not generate a binary version of this policy as you did in the [Create code integrity policies from golden PCs](#create-code-integrity-policies-from-golden-pcs) section. This is because code integrity policies created from an audit log are not intended to run as stand-alone policies but rather to update existing code integrity policies.
### <a href="" id="merge-code-integrity"></a>
**Merge code integrity policies**
### Merge code integrity policies
When you develop code integrity policies, you will occasionally need to merge two policies. A common example is when a code integrity policy is initially created and audited. Another example is when you create a single master policy by using multiple code integrity policies previously created from golden PCs. Because each Windows 10 machine can have only one code integrity policy, it is important to properly maintain these policies. In this example, audit events have been saved into a secondary code integrity policy that you then merge with the initial code integrity policy.
**Note**
The following example uses the code integrity policy .xml files that you created in the [Create code integrity policies from golden PCs](#create-code-golden) and [Audit code integrity policies](#audit-code-integrity) sections. You can follow this process, however, with any two code integrity policies you would like to combine.
The following example uses the code integrity policy .xml files that you created in the [Create code integrity policies from golden PCs](#create-code-integrity-policies-from-golden-pcs) and [Audit code integrity policies](#audit-code-integrity-policies) sections. You can follow this process, however, with any two code integrity policies you would like to combine.
To merge two code integrity policies, complete the following steps in an elevated Windows PowerShell session:
@ -1207,8 +1055,6 @@ Every code integrity policy is created with audit mode enabled. After you have s
**Note**
Every code integrity policy should be tested in audit mode first. For information about how to audit code integrity policies, see the [Audit code integrity policies](#audit-code-integrity) section.
1. Initialize the variables that will be used:
`$CIPolicyPath=$env:userprofile+"\Desktop\"`
@ -1246,13 +1092,13 @@ Every code integrity policy should be tested in audit mode first. For informatio
Now that this policy has been enforced, you can deploy it to your test machines. Rename the policy to SIPolicy.p7b and copy it to C:\\Windows\\System32\\CodeIntegrity for testing, or deploy the policy through Group Policy by following the instructions in the [Deploy and manage code integrity policies with Group Policy](#deploy-manage-code-gp) section, or through client management software by following the instructions in the section “Deploying and managing code integrity policies by using Microsoft client management solutions.”
Now that this policy has been enforced, you can deploy it to your test machines. Rename the policy to SIPolicy.p7b and copy it to C:\\Windows\\System32\\CodeIntegrity for testing, or deploy the policy through Group Policy by following the instructions in the [Deploy and manage code integrity policies with Group Policy](#deploy-and-manage-code-integrity-policies-with-group-policy) section, or through client management software by following the instructions in the section “Deploying and managing code integrity policies by using Microsoft client management solutions.”
**Signing code integrity policies with SignTool.exe**
Signed code integrity policies give organizations the highest level of malware protection available in Windows 10. In addition to their enforced policy rules, signed policies cannot be modified or deleted by a user or administrator on the machine. These policies are designed to prevent administrative tampering and kernel mode exploit access. With this in mind, it is much more difficult to remove signed code integrity policies than unsigned ones. Before you sign and deploy a signed code integrity policy, Microsoft recommends that you audit the policy to discover any blocked applications that should be allowed to run. For more information about how to audit code integrity policies, see the [Audit code integrity policies](#audit-code-integrity) section.
Signed code integrity policies give organizations the highest level of malware protection available in Windows 10. In addition to their enforced policy rules, signed policies cannot be modified or deleted by a user or administrator on the machine. These policies are designed to prevent administrative tampering and kernel mode exploit access. With this in mind, it is much more difficult to remove signed code integrity policies than unsigned ones. Before you sign and deploy a signed code integrity policy, Microsoft recommends that you audit the policy to discover any blocked applications that should be allowed to run. For more information about how to audit code integrity policies, see the [Audit code integrity policies](#audit-code-integrity-policies) section.
Signing code integrity policies by using an on-premises CA-generated certificate or a purchased code signing certificate is straightforward. If you do not currently have a code signing certificate exported in .pfx format (containing private keys, extensions, and root certificates), see [Create a Device Guard code signing certificate](#create-dg-code) to create one with your on-premises CA. Before signing code integrity policies for the first time, be sure to enable rule options 9 and 10 to leave troubleshooting options available to test administrators. When validated and ready for enterprise deployment, you can remove these options. For information about how to add rule options, see the [Code integrity policy rules](#code-integrity-policy-rules) section.
Signing code integrity policies by using an on-premises CA-generated certificate or a purchased code signing certificate is straightforward. If you do not currently have a code signing certificate exported in .pfx format (containing private keys, extensions, and root certificates), see [Create a Device Guard code signing certificate](#create-a-device-guard-code-signing-certificate) to create one with your on-premises CA. Before signing code integrity policies for the first time, be sure to enable rule options 9 and 10 to leave troubleshooting options available to test administrators. When validated and ready for enterprise deployment, you can remove these options. For information about how to add rule options, see the [Code integrity policy rules](#code-integrity-policy-rules) section.
**Note**
Signing code integrity policies is the last step in a code integrity deployment. It is much more difficult to remove a signed code integrity policy than an unsigned one. Before you deploy a signed code integrity policy to deployed client computers, be sure to test its effect on a subset of machines.
@ -1265,8 +1111,6 @@ To sign a code integrity policy with SignTool.exe, you need the following compon
- An internal CA code signing certificate or a purchased code signing certificate
If you do not have a code signing certificate, see the [Create a Device Guard code signing certificate](#create-dg-code) section for instructions on how to create one. If you use an alternate certificate or code integrity policy, be sure to update the following steps with the appropriate variables and certificate so that the commands will function properly. To sign the existing code integrity policy, copy each of the following commands into an elevated Windows PowerShell session:
1. Initialize the variables that will be used:
@ -1276,8 +1120,6 @@ If you do not have a code signing certificate, see the [Create a Device Guard co
**Note**
This example uses the code integrity policy that you created in the [Create code integrity policies from golden PCs](#create-code-golden) section. If you are signing another policy, be sure to update the **$CIPolicyPath** and **$CIPolicyBin** variables with the correct information.
2. Import the .pfx code signing certificate. Import the code signing certificate that you will use to sign the code integrity policy into the signing user’s personal store on the machine that will be doing the signing. In this example, you use the certificate that was created in the [Create a Device Guard code signing certificate](#create-dg-code) section.
3. Export the .cer code signing certificate. After the code signing certificate has been imported, export the .cer version to your desktop. This version will be added to the policy so that it can be updated later.
@ -1290,15 +1132,11 @@ If you do not have a code signing certificate, see the [Create a Device Guard co
*<Path to exported .cer certificate>* should be the full path to the certificate that you exported in step 3.
**Note**
Adding update signers is crucial to being able to modify or disable this policy in the future. For more information about how to disable signed code integrity policies, see the [Disable signed code integrity policies within Windows](#disable-signed-code) section.
**Note**
*<Path to exported .cer certificate>* should be the full path to the certificate that you exported in step 3.
**Note**
Adding update signers is crucial to being able to modify or disable this policy in the future. For more information about how to disable signed code integrity policies, see the [Disable signed code integrity policies within Windows](#disable-signed-code) section.
6. Remove the unsigned policy rule option:
@ -1315,13 +1153,9 @@ If you do not have a code signing certificate, see the [Create a Device Guard co
**Note**
The *<Path to signtool.exe>* variable should be the full path to the SignTool.exe utility. **ContosoDGSigningCert** is the subject name of the certificate that will be used to sign the code integrity policy. You should import this certificate to your personal certificate store on the machine you use to sign the policy.
9. Validate the signed file. When complete, the commands should output a signed policy file called DeviceGuardPolicy.bin.p7 to your desktop. You can deploy this file the same way you deploy an enforced or non-enforced policy. For information about how to deploy code integrity policies, see the [Deploy and manage code integrity policies with Group Policy](#deploy-and-manage-code-integrity-policies-with-group-policy) section.
9. Validate the signed file. When complete, the commands should output a signed policy file called DeviceGuardPolicy.bin.p7 to your desktop. You can deploy this file the same way you deploy an enforced or non-enforced policy. For information about how to deploy code integrity policies, see the [Deploy and manage code integrity policies with Group Policy](#deploy-manage-code-gp) section.
### <a href="" id="disable-unsigned-code"></a>
**Disable unsigned code integrity policies**
### Disable unsigned code integrity policies
There may come a time when an administrator wants to disable a code integrity policy. For unsigned code integrity policies, this process is simple. Depending on how the code integrity policy was deployed, unsigned policies can be disabled in one of two ways. If a code integrity policy was manually enabled and copied to the code integrity folder location, simply delete the file and restart the machine. The following locations can contain executing code integrity policies:
@ -1331,9 +1165,7 @@ There may come a time when an administrator wants to disable a code integrity po
If the code integrity policy was deployed by using Group Policy, the GPO that is currently enabling and deploying the policy must be set to disabled. Then, the code integrity policy will be disabled on the next computer restart.
### <a href="" id="disable-signed-code"></a>
**Disable signed code integrity policies within Windows**
### Disablesignedcode integrity policies within Windows
Signed policies protect Windows from administrative manipulation as well as malware that has gained administrative-level access to the system. For this reason, signed code integrity policies are intentionally more difficult to remove than unsigned policies. They inherently protect themselves from modification or removal and therefore are difficult even for administrators to remove successfully. If the signed code integrity policy is manually enabled and copied to the CodeIntegrity folder, to remove the policy, you must complete the following steps:
@ -1344,15 +1176,12 @@ For reference, signed code integrity policies should be replaced and removed fro
1. Replace the existing policy with another signed policy that has the **6 Enabled: Unsigned System Integrity Policy** rule option enabled.
**Note**
To take effect, this policy must be signed with a certificate previously added to the **UpdatePolicySigners** section of the original signed policy you want to replace.
2. Restart the client computer.
3. Verify that the new signed policy exists on the client.
@ -1360,8 +1189,6 @@ For reference, signed code integrity policies should be replaced and removed fro
**Note**
If the signed policy that contains rule option 6 has not been processed on the client, the addition of an unsigned policy may cause boot failures.
4. Delete the new policy.
5. Restart the client computer.
@ -1382,17 +1209,13 @@ If the signed code integrity policy has been deployed using by using Group Polic
**Note**
If the signed policy that contains rule option 6 has not been processed on the client, the addition of an unsigned policy may cause boot failures.
4. Set the GPO to disabled.
5. Delete the new policy.
6. Restart the client computer.
### <a href="" id="disable-signed-code-bios"></a>
**Disable signed code integrity policies within the BIOS**
### Disablesignedcode integrity policies within the BIOS
There may be a time when signed code integrity policies cause a boot failure. Because code integrity policies enforce kernel mode drivers, it is important that they be thoroughly tested on each software and hardware configuration before being enforced and signed. Signed code integrity policies are validated in the pre-boot sequence by using Secure Boot. When you disable the Secure Boot feature in the BIOS, and then delete the file from the following locations on the operating system disk, it allows the system to boot into Windows:
@ -1407,15 +1230,11 @@ There may be a time when signed code integrity policies cause a boot failure. Be
Code integrity policies can easily be deployed and managed with Group Policy. A Device Guard administrative template will be available in Windows Server 2016 that allows you to simplify deployment of Device Guard hardware-based security features and code integrity policies. The following procedure walks you through how to deploy a code integrity policy called **DeviceGuardPolicy.bin** to a test OU called *DG Enabled PCs* by using a GPO called **Contoso GPO Test**.
**Note**
This walkthrough requires that you have previously created a code integrity policy and have a Windows 10 client PC on which to test a Group Policy deployment. For more information about how to create a code integrity policy, see the [Create code integrity polices from golden PCs](#create-code-golden) section.
This walkthrough requires that you have previously created a code integrity policy and have a Windows 10 client PC on which to test a Group Policy deployment. For more information about how to create a code integrity policy, see the [Create code integrity polices from golden PCs](#create-code-integrity-polices-from-golden-pcs) section.
**Note**
Signed code integrity policies can cause boot failures when deployed. Microsoft recommends that signed code integrity policies be thoroughly tested on each hardware platform before enterprise deployment.
To deploy and manage a code integrity policy with Group Policy:
1. On a domain controller on a client computer on which RSAT is installed, open the GPMC by running **GPMC.MSC** or searching for “Group Policy Management” in Windows Search.
@ -1425,11 +1244,9 @@ To deploy and manage a code integrity policy with Group Policy:
**Note**
The DG Enabled PCs OU is just an example of where to link the test GPO created in this section. Any OU name can be used. Also, security group filtering is an option when considering policy partitioning options based on the strategy discussed in the [Approach enterprise code integrity deployment](#approach-enterprise) section.
Figure 24. Create a GPO
Figure 24. Create a GPO
3. Name new GPO **Contoso GPO Test**. This example uses Contoso GPO Test as the name of the GPO. You can choose any name that you prefer for this example.
@ -1443,26 +1260,21 @@ To deploy and manage a code integrity policy with Group Policy:
6. In the **Display Code Integrity Policy** dialog box, select the **Enabled** option, and then specify the code integrity policy deployment path.
In this policy setting, you specify either the local path in which the policy will exist on the client computer or a Universal Naming Convention (UNC) path that the client computers will look to retrieve the latest version of the policy. This example copied the DeviceGuardPolicy.bin file onto the test machine and will enable this setting and use the file path C:\\Windows\\System32\\CodeIntegrity\\DeviceGuardPolicy.bin, as shown in Figure 26.
In this policy setting, you specify either the local path in which the policy will exist on the client computer or a Universal Naming Convention (UNC) path that the client computers will look to retrieve the latest version of the policy. This example copied the DeviceGuardPolicy.bin file onto the test machine and will enable this setting and use the file path C:\\Windows\\System32\\CodeIntegrity\\DeviceGuardPolicy.bin, as shown in Figure 26.
**Note**
*DeviceGuardPolicy.bin* is not a required policy name: It was simply used in the [Create code integrity policies from golden PCs](#create-code-golden) section and so is used here, as well. Also, this policy file does not need to be copied to every computer. Alternatively, you can copy the code integrity policies to a file share to which the computer accounts have access. Any policy selected here is converted to SIPolicy.p7b when it is deployed to the individual client computers.
**Note**
*DeviceGuardPolicy.bin* is not a required policy name: It was simply used in the [Create code integrity policies from golden PCs](#create-code-golden) section and so is used here, as well. Also, this policy file does not need to be copied to every computer. Alternatively, you can copy the code integrity policies to a file share to which the computer accounts have access. Any policy selected here is converted to SIPolicy.p7b when it is deployed to the individual client computers.
Figure 26. Enable the code integrity policy
Figure 26. Enable the code integrity policy
**Note**
You may have noticed that the GPO setting references a .p7b file and this example uses a .bin file for the policy. Regardless of the type of policy you deploy (.bin, .p7b, or .p7), they are all converted to SIPolicy.p7b when dropped on the Windows 10 client computers. Make your code integrity policies friendly and allow the system to convert the policy names for you to ensure that the policies are easily distinguishable when viewed in a share or any other central repository.
**Note**
You may have noticed that the GPO setting references a .p7b file and this example uses a .bin file for the policy. Regardless of the type of policy you deploy (.bin, .p7b, or .p7), they are all converted to SIPolicy.p7b when dropped on the Windows 10 client computers. Make your code integrity policies friendly and allow the system to convert the policy names for you to ensure that the policies are easily distinguishable when viewed in a share or any other central repository.
7. Close the Group Policy Management Editor, and then restart the Windows 10 test machine. Restarting the client computer updates the code integrity policy. For information about how to audit code integrity policies, see the [Audit code integrity policies](#audit-code-integrity)section.
## <a href="" id="create-dg-code"></a>Create a Device Guard code signing certificate
7. Close the Group Policy Management Editor, and then restart the Windows 10 test machine. Restarting the client computer updates the code integrity policy. For information about how to audit code integrity policies, see the [Audit code integrity policies](#audit-code-integrity-policies)section.
## Create a Device Guard code signing certificate
To sign catalog files or code integrity policies internally, you will either need a publicly issued code signing certificate or an internal CA. If you have purchased a code signing certificate, you can skip these steps and proceed to the sections that outline the steps to sign catalog files and code integrity policies. If you have not purchased a certificate but have an internal CA, complete these steps to create a code signing certificate:
@ -1531,8 +1343,6 @@ Now that the template is available to be issued, you must request one from the W
**Note**
If a certificate manager is required to approve any issued certificates and you selected to require management approval on the template, the request will need to be approved in the CA before it will be issued to the client.
This certificate must be installed in the user’s personal store on the computer that will be signing the catalog files and code integrity policies. If the signing is going to be taking place on the machine on which you just requested the certificate, exporting the certificate to a .pfx file will not be required because it already exists in your personal store. If you are signing on another computer, you will need to export the .pfx certificate with the necessary keys and properties. To do so, complete the following steps:
1. Right-click the certificate, point to **All Tasks**, and then click **Export**.
@ -1548,23 +1358,12 @@ When the certificate has been exported, import it into the personal store for th
@ -62,7 +62,7 @@ Learn about keeping Windows10 and Windows10 Mobile secure.
</tr>
<trclass="odd">
<tdalign="left"><p>[Protect your enterprise data using enterprise data protection (EDP)](protect-enterprise-data-using-edp.md)</p></td>
<tdalign="left"><p>With the increase of employee-owned devices in the enterprise, there’s also an increasing risk of accidental data leak through apps and services, like email, social media, and the public cloud, which are outside of the enterprise’s control. For example, when an employee sends the latest engineering pictures to their personal email account, copies and pastes product info to a public Yammer group or tweet, or saves an in-progress sales report to their public cloud storage.</p></td>
<tdalign="left"><p>With the increase of employee-owned devices in the enterprise, there’s also an increasing risk of accidental data leak through apps and services, like email, social media, and the public cloud, which are outside of the enterprise’s control. For example, when an employee sends the latest engineering pictures from their personal email account, copies and pastes product info into a tweet, or saves an in-progress sales report to their public cloud storage.</p></td>
</tr>
<trclass="even">
<tdalign="left"><p>[Use Windows Event Forwarding to help with intrusion detection](use-windows-event-forwarding-to-assist-in-instrusion-detection.md)</p></td>
@ -33,7 +33,7 @@ When a user encounters an error when creating the work PIN, advise the user to t
1. Try to create the PIN again. Some errors are transient and resolve themselves.
2.Log out, log in, and try to create the PIN again.
2.Sign out, sign in, and try to create the PIN again.
3. Reboot the device and then try to create the PIN again.
@ -44,11 +44,7 @@ When a user encounters an error when creating the work PIN, advise the user to t
If the error occurs again, check the error code against the following table to see if there is another mitigation for that error. When no mitigation is listed in the table, contact Microsoft Support for assistance.
<table>
<colgroup>
<colwidth="33%"/>
<colwidth="33%"/>
<colwidth="33%"/>
</colgroup>
<thead>
<trclass="header">
<thalign="left">Hex</th>
@ -57,20 +53,13 @@ If the error occurs again, check the error code against the following table to s
</tr>
</thead>
<tbody>
<trclass="odd">
<tdalign="left">0x801C03ED</td>
<tdalign="left"><p>Multi-factor authentication is required for a 'ProvisionKey' operation, but was not performed</p>
<p>-or-</p>
<p>Token was not found in the Authorization header</p>
<p>-or-</p>
<p>Failed to read one or more objects</p></td>
<tdalign="left">Unjoin the device from Azure Active Directory (Azure AD) and rejoin</td>
</tr>
<trclass="even">
<tdalign="left">0x801C044D</td>
<tdalign="left">Authorization token does not contain device ID</td>
<tdalign="left">Unjoin the device from Azure AD and rejoin</td>
</tr>
<trclass="odd">
<tdalign="left">0x80090036</td>
<tdalign="left">User cancelled an interactive dialog</td>
@ -95,6 +84,10 @@ If the error occurs again, check the error code against the following table to s
<tdalign="left">0x80090005</td>
<tdalign="left">NTE_BAD_DATA</td>
<tdalign="left">Unjoin the device from Azure AD and rejoin</td>
</tr><trclass="even">
<tdalign="left">0x80090029</td>
<tdalign="left">TPM is not set up.</td>
<tdalign="left">Sign on with an administrator account. Click **Start**, type "tpm.msc", and select **tpm.msc Microsoft Common Console Document**. In the **Actions** pane, select **Prepare the TPM**. </td>
</tr>
<trclass="even">
<tdalign="left">0x80090031</td>
@ -124,17 +117,17 @@ If the error occurs again, check the error code against the following table to s
<trclass="odd">
<tdalign="left">0x801C0010</td>
<tdalign="left">The AIK certificate is not valid or trusted</td>
<tdalign="left">Log out and then log in again.</td>
<tdalign="left">Sign out and then sign in again.</td>
</tr>
<trclass="even">
<tdalign="left">0x801C0011</td>
<tdalign="left">The attestation statement of the transport key is invalid</td>
<tdalign="left">Log out and then log in again.</td>
<tdalign="left">Sign out and then sign in again.</td>
</tr>
<trclass="odd">
<tdalign="left">0x801C0012</td>
<tdalign="left">Discovery request is not in a valid format</td>
<tdalign="left">Log out and then log in again.</td>
<tdalign="left">Sign out and then sign in again.</td>
</tr>
<trclass="even">
<tdalign="left">0x801C0015</td>
@ -159,7 +152,7 @@ If the error occurs again, check the error code against the following table to s
<trclass="even">
<tdalign="left">0x801C03E9</td>
<tdalign="left">Server response message is invalid</td>
<tdalign="left">Log out and then log in again.</td>
<tdalign="left">Sign out and then sign in again.</td>
</tr>
<trclass="odd">
<tdalign="left">0x801C03EA</td>
@ -169,37 +162,42 @@ If the error occurs again, check the error code against the following table to s
<trclass="even">
<tdalign="left">0x801C03EB</td>
<tdalign="left">Server response http status is not valid</td>
<tdalign="left">Log out and then log in again.</td>
<tdalign="left">Sign out and then sign in again.</td>
</tr>
<trclass="odd">
<tdalign="left">0x801C03EC</td>
<tdalign="left">Unhandled exception from server.</td>
<tdalign="left">Log out and then log in again.</td>
<tdalign="left">sign out and then sign in again.</td>
</tr>
<trclass="even">
<tdalign="left">0x801C03ED</td>
<tdalign="left">The request sent to the server was invalid.</td>
<tdalign="left">Log out and then log in again.</td>
<tdalign="left"><p>Multi-factor authentication is required for a 'ProvisionKey' operation, but was not performed</p>
<p>-or-</p>
<p>Token was not found in the Authorization header</p>
<p>-or-</p>
<p>Failed to read one or more objects</p>
<p>-or-</p><p>The request sent to the server was invalid.</p></td>
<tdalign="left">Sign out and then sign in again. If that doesn't resolve the issue, unjoin the device from Azure Active Directory (Azure AD) and rejoin.</td>
</tr>
<trclass="odd">
<tdalign="left">0x801C03EE</td>
<tdalign="left">Attestation failed</td>
<tdalign="left">Log out and then log in again.</td>
<tdalign="left">Sign out and then sign in again.</td>
</tr>
<trclass="even">
<tdalign="left">0x801C03EF</td>
<tdalign="left">The AIK certificate is no longer valid</td>
<tdalign="left">Log out and then log in again.</td>
<tdalign="left">Sign out and then sign in again.</td>
</tr>
<trclass="odd">
<tdalign="left">0x801C044D</td>
<tdalign="left">Unable to obtain user token</td>
<tdalign="left">Log out and then log in again. Check network and credentials.</td>
<tdalign="left">Sign out and then sign in again. Check network and credentials.</td>
</tr>
<trclass="even">
<tdalign="left">0x801C044E</td>
<tdalign="left">Failed to receive user creds input</td>
<tdalign="left">Log out and then log in again.</td>
<tdalign="left">Sign out and then sign in again.</td>
</tr>
</tbody>
</table>
@ -214,6 +212,7 @@ For errors listed in this table, contact Microsoft Support for assistance.
@ -405,7 +406,7 @@ Table 1. Deployment requirements for Microsoft Passport
Note that the current release of Windows10 supports the Azure AD–only scenarios. Microsoft provides the forward-looking guidance in Table 1 to help organizations prepare their environments for planned future releases of Microsoft Passport for Work capabilities.
Note that the current release of Windows10 supports the Azure AD–only (RTM) and hybrid scenarios (RTM + November Update). Microsoft provides the forward-looking guidance in Table 1 to help organizations prepare their environments for planned future releases of Microsoft Passport for Work capabilities.
**Select policy settings**
@ -465,17 +466,19 @@ In the Windows10 initial release, Microsoft supports the following Microsoft P
- Microsoft Passport for Work support for organizations that have cloud-only Azure AD deployments
- Group Policy settings to control Microsoft Passport PIN length and complexity
- Group Policy and MDM settings to control Microsoft Passport PIN length and complexity
In the November 2015 release, Microsoft supports the following Microsoft Passport and Windows Hello features:
- Key-based Microsoft Passport for Work credentials for on-premises Azure AD deployments and hybrid on-premises/Azure AD deployments
- Microsoft Passport for Work certificates issued by a trusted PKI, including smart card and virtual smart card certificates
In future releases of Windows10, we plan to add support for additional features:
- Additional biometric identifier types, including iris recognition
- Key-based Microsoft Passport for Work credentials for on-premises Azure AD deployments and hybrid on-premises/Azure AD deployments
- Microsoft Passport for Work certificates issued by a trusted PKI, including smart card and virtual smart card certificates
- TPM attestation to protect keys so that a malicious user or program can’t create keys in software (because those keys won’t be TPM attested and can thus be identified as fake)
-Key-based and certificate-based Microsoft Passport for Work credentials for on-premises AD deployments
-TPM attestation to protect keys so that a malicious user or program can’t create keys in software (because those keys won’t be TPM attested and can thus be identified as fake)
In the longer term, Microsoft will continue to improve on and expand the features of both Microsoft Passport and Windows Hello to cover additional customer requirements for manageability and security. We also are working with the FIDO Alliance and a variety of third parties to encourage adoption of Microsoft Passport by both web and LOB application developers.
<spanstyle="color:#ED1C24;">[Some information relates to pre-released product, which may be substantially modified before it's commercially released. Microsoft makes no warranties, express or implied, with respect to the information provided here.]</span>
With the increase of employee-owned devices in the enterprise, there’s also an increasing risk of accidental data leak through apps and services, like email, social media, and the public cloud, which are outside of the enterprise’s control. For example, when an employee sends the latest engineering pictures to their personal email account, copies and pastes product info to a public Yammer group or tweet, or saves an in-progress sales report to their public cloud storage.
With the increase of employee-owned devices in the enterprise, there’s also an increasing risk of accidental data leak through apps and services, like email, social media, and the public cloud, which are outside of the enterprise’s control. For example, when an employee sends the latest engineering pictures from their personal email account, copies and pastes product info into a tweet, or saves an in-progress sales report to their public cloud storage.
Enterprise data protection (EDP) helps to protect against this potential data leakage without otherwise interfering with the employee experience. EDP also helps to protect enterprise apps and data against accidental data leak on enterprise-owned devices and personal devices that employees bring to work without requiring changes to your environment or other apps. Finally, another data protection technology, Azure Rights Management also works alongside EDP to extend data protection for data that leaves the device, such as when email attachments are sent from an enterprise aware version of a rights management mail client.
@ -31,7 +31,15 @@ Trusted Platform Module (TPM) technology is designed to provide hardware-based,
The most common TPM functions are used for system integrity measurements and for key creation and use. During the boot process of a system, the boot code that is loaded (including firmware and the operating system components) can be measured and recorded in the TPM. The integrity measurements can be used as evidence for how a system started and to make sure that a TPM-based key was used only when the correct software was used to boot the system.
Different versions of the TPM are defined in specifications by the Trusted Computing Group (TCG).
Traditionally, TPMs have been discrete chips soldered to a computer’s motherboard. Such implementations allow the computer’s original equipment manufacturer (OEM) to evaluate and certify the TPM separate from the rest of the system. Although discrete TPM implementations are still common, they can be problematic for integrated devices that are small or have low power consumption. Some newer TPM implementations integrate TPM functionality into the same chipset as other platform components while still providing logical separation similar to discrete TPM chips.
TPMs are passive: they receive commands and return responses. To realize the full benefit of a TPM, the OEM must carefully integrate system hardware and firmware with the TPM to send it commands and react to its responses. TPMs were originally designed to provide security and privacy benefits to a platform’s owner and users, but newer versions can provide security and privacy benefits to the system hardware itself. Before it can be used for advanced scenarios, however, a TPM must be provisioned. Windows 10 automatically provisions a TPM, but if the user reinstalls the operating system, he or she may need to tell the operating system to explicitly provision the TPM again before it can use all the TPM’s features.
The Trusted Computing Group (TCG) is the nonprofit organization that publishes and maintains the TPM specification. The TCG exists to develop, define, and promote vendor-neutral, global industry standards that support a hardware-based root of trust for interoperable trusted computing platforms. The TCG also publishes the TPM specification as the international standard ISO/IEC 11889, using the Publicly Available Specification Submission Process that the Joint Technical Committee 1 defines between the International Organization for Standardization (ISO) and the International Electrotechnical Commission (IEC).
OEMs implement the TPM as a component in a trusted computing platform, such as a PC, tablet, or phone. Trusted computing platforms use the TPM to support privacy and security scenarios that software alone cannot achieve. For example, software alone cannot reliably report whether malware is present during the system startup process. The close integration between TPM and platform increases the transparency of the startup process and supports evaluating device health by enabling reliable measuring and reporting of the software that starts the device. Implementation of a TPM as part of a trusted computing platform provides a hardware root of trust—that is, it behaves in a trusted way. For example, if a key stored in a TPM has properties that disallow exporting the key, that key truly cannot leave the TPM.
The TCG designed the TPM as a low-cost, mass-market security solution that addresses the requirements of different customer segments. There are variations in the security properties of different TPM implementations just as there are variations in customer and regulatory requirements for different sectors. In public-sector procurement, for example, some governments have clearly defined security requirements for TPMs whereas others do not.
**Note**
Some information relates to pre-released product which may be substantially modified before it's commercially released. Microsoft makes no warranties, express or implied, with respect to the information provided here.
@ -41,11 +49,10 @@ Some information relates to pre-released product which may be substantially modi
## TPM 1.2 vs. 2.0 comparison
From an industry standard, Microsoft has been an industry leader in moving and standardizing on TPM 2.0. As indicated in the table below, TPM 2.0 has many key realized benefits across algorithms, crypto, hierarchy, root keys, authorization and NV RAM.
From an industry standard, Microsoft has been an industry leader in moving and standardizing on TPM 2.0, which has many key realized benefits across algorithms, crypto, hierarchy, root keys, authorization and NV RAM.
## Why TPM 2.0?
TPM 2.0 products and systems have important security advantages over TPM 1.2, including:
- The TPM 1.2 spec only allows for the use of RSA and the SHA-1 hashing algorithm.
@ -65,7 +72,6 @@ TPM 2.0 products and systems have important security advantages over TPM 1.2, in
## Discrete or firmware TPM?
Windows uses discrete and firmware TPM in the same way. Windows gains no functional advantage or disadvantage from either option.
From a security standpoint, discrete and firmware share the same characteristics;
@ -77,20 +83,22 @@ From a security standpoint, discrete and firmware share the same characteristics
For more info, see [fTPM: A Firmware-based TPM 2.0 Implementation](http://research.microsoft.com/apps/pubs/?id=258236).
## TPM 2.0 Compliance for Windows 10 in the future
## Is there any importance for TPM for consumer?
For end consumers, TPM is behind the scenes but still very relevant for Hello, Passport and in the future, many other key features in Windows 10. It offers the best Passport experience, helps encrypt passwords, secures streaming high quality 4K content and builds on our overall Windows 10 experience story for security as a critical pillar. Using Windows on a system with a TPM enables a deeper and broader level of security coverage.
All shipping devices for Windows 10 across all SKU types must be using TPM 2.0 discrete or firmware from **July 28, 2016**. This requirement will be enforced through our Windows Hardware Certification program.
## TPM 2.0 Compliance for Windows 10
### Windows10 for desktop editions (Home, Pro, Enterprise, and Education)
- With Windows 10 as with Windows 8, all connected standby systems are required to include TPM 2.0 support.
- For Windows 10 and later, if a SoC is chosen that includes an integrated fTPM2.0, the device must ship with the fTPM FW support or a discrete TPM 1.2 or 2.0.
- Starting **July 28th, 2016** all devices shipping with Windows 10 desktop must implement TPM 2.0 and ship with the TPM enabled.
-As of July 28, 2016, all new device models, lines or series (or if you are updating the hardware configuration of a existing model, line or series with a major update, such as CPU, graphic cards) must implement and enable by default TPM 2.0 (details in section 3.7, https://msdn.microsoft.com/library/windows/hardware/dn915086(v=vs.85).aspx)
## Two implementation options:
• Discrete TPM chip as a separate discrete component
• Firmware TPM solution using Intel PTT (platform trust technology) or AMD
### Windows10 Mobile
- All devices shipping with Windows10 Mobile must implement TPM 2.0 and ship with the TPM enabled.
- All devices shipping with Windows10 Mobile must implement TPM 2.0 and ship with the TPM 2.0 enabled.
### IoT Core
@ -102,7 +110,6 @@ All shipping devices for Windows 10 across all SKU types must be using TPM 2.0 d
## TPM and Windows Features
The following table defines which Windows features require TPM support. Some features are not applicable to Windows 7/8/8.1 and are noted accordingly.
<table>
@ -124,7 +131,7 @@ The following table defines which Windows features require TPM support. Some fea
</thead>
<tbody>
<trclass="odd">
<tdalign="left">Measure Boot</td>
<tdalign="left">Measured Boot</td>
<tdalign="left">Required</td>
<tdalign="left">Required</td>
<tdalign="left">Required</td>
@ -147,7 +154,7 @@ The following table defines which Windows features require TPM support. Some fea
<trclass="even">
<tdalign="left">Passport: MSA or Local Account</td>
<tdalign="left">n/a</td>
<tdalign="left">Not Required</td>
<tdalign="left">Required</td>
<tdalign="left">Required</td>
<tdalign="left">TPM 2.0 is required with HMAC and EK certificate for key attestation support.</td>
</tr>
@ -175,7 +182,7 @@ The following table defines which Windows features require TPM support. Some fea
<trclass="even">
<tdalign="left">Device Health Attestation</td>
<tdalign="left">n/a</td>
<tdalign="left">Not Required</td>
<tdalign="left">Required</td>
<tdalign="left">Required</td>
<tdalign="left"></td>
</tr>
@ -240,6 +247,7 @@ There are a variety of TPM manufacturers for both discrete and firmware.
<tdalign="left"><ul>
<li>Infineon</li>
<li>Nuvoton</li>
<li>Atmel</li>
<li>NationZ</li>
<li>ST Micro</li>
</ul></td>
@ -274,11 +282,12 @@ There are a variety of TPM manufacturers for both discrete and firmware.
<trclass="even">
<tdalign="left">Intel</td>
<tdalign="left"><ul>
<li>Clovertrail</li>
<li>Haswell</li>
<li>Broadwell</li>
<li>Skylake</li>
<li>Atom (CloverTrail)
<li>Baytrail</li>
<li>4th generation(Haswell)</li>
<li>5th generation(Broadwell)</li>
<li>Braswell</li>
<li>Skylake</li>
</ul></td>
</tr>
<trclass="odd">
@ -301,7 +310,7 @@ There are a variety of TPM manufacturers for both discrete and firmware.
### Certified TPM parts
Government customers and enterprise customers in regulated industries may have acquisition standards that require use of common certified TPM parts. As a result, OEMs, who provide the devices, may be required to use only certified TPM components on their commercial class systems. Discrete TPM 2.0 vendors have targeted completion of certification by the end of 2015.
Government customers and enterprise customers in regulated industries may have acquisition standards that require use of common certified TPM parts. As a result, OEMs, who provide the devices, may be required to use only certified TPM components on their commercial class systems. Discrete TPM 2.0 vendors have completion certification.
title: User Account Control Group Policy and registry key settings (Windows 10)
description: Here's a list of UAC Group Policy and registry key settings that your organization can use to manage UAC.
ms.prod: W10
ms.mktglfcycl: deploy
ms.sitesec: library
---
# User Account Control Group Policy and registry key settings
**Applies to**
- Windows 10
- Windows Server 2016 Technical Preview
## Group Policy settings
There are 10 Group Policy settings that can be configured for User Account Control (UAC). The table lists the default for each of the policy settings, and the following sections explain the different UAC policy settings and provide recommendations. These policy settings are located in **Security Settings\\Local Policies\\Security Options** in the Local Security Policy snap-in. For more information about each of the Group Policy settings, see the Group Policy description. For information about the registry key settings, see [Registry key settings](#registry-key-settings).
| Group Policy setting | Registry key | Default |
| - | - | - | - |
| [User Account Control: Admin Approval Mode for the built-in Administrator account](#user-account-control-admin-approval-mode-for-the-built-in-administrator-account) | FilterAdministratorToken | Disabled |
| [User Account Control: Allow UIAccess applications to prompt for elevation without using the secure desktop](#user-account-control-allow-uiaccess-applications-to prompt-for-elevation-without-using-the-secure-desktop) | EnableUIADesktopToggle | Disabled |
| [User Account Control: Behavior of the elevation prompt for administrators in Admin Approval Mode](#user-account-control-behavior-of-the-elevation-prompt-for-administrators-in-admin-approval-mode) | ConsentPromptBehaviorAdmin | Prompt for consent for non-Windows binaries |
| [User Account Control: Behavior of the elevation prompt for standard users](#user-account-control-behavior-of-the-elevation-prompt-for-standard-users) | ConsentPromptBehaviorUser | Prompt for credentials on the secure desktop |
| [User Account Control: Detect application installations and prompt for elevation](#user-account-control-detect-application-installations-and-prompt-for-elevation) | EnableInstallerDetection | Enabled (default for home)<br/>Disabled (default for enterprise) |
| [User Account Control: Only elevate executables that are signed and validated](#user-account-control-only-elevate-executables-that-are-signed-and-validated) | ValidateAdminCodeSignatures | Disabled |
| [User Account Control: Only elevate UIAccess applications that are installed in secure locations](#user-account-control-only-elevate-uiaccess-applications-that-are-installed-in-secure-locations) | EnableSecureUIAPaths | Enabled |
| [User Account Control: Run all administrators in Admin Approval Mode](#user-account-control-run-all-administrators-in-admin-approval-mode) | EnableLUA | Enabled |
| [User Account Control: Switch to the secure desktop when prompting for elevation](#user-account-control-switch-to-the-secure-desktop-when-prompting-for-elevation) | PromptOnSecureDesktop | Enabled |
| [User Account Control: Virtualize file and registry write failures to per-user locations](#user-account-control-virtualize-file-and-registry-write-failures-to-per-user-locations) | EnableVirtualization | Enabled |
### User Account Control: Admin Approval Mode for the built-in Administrator account
The **User Account Control: Admin Approval Mode for the built-in Administrator account** policy setting controls the behavior of Admin Approval Mode for the built-in Administrator account.
The options are:
-**Enabled.** The built-in Administrator account uses Admin Approval Mode. By default, any operation that requires elevation of privilege will prompt the user to approve the operation.
-**Disabled.** (Default) The built-in Administrator account runs all applications with full administrative privilege.
### User Account Control: Allow UIAccess applications to prompt for elevation without using the secure desktop
The **User Account Control: Allow UIAccess applications to prompt for elevation without using the secure desktop** policy setting controls whether User Interface Accessibility (UIAccess or UIA) programs can automatically disable the secure desktop for elevation prompts used by a standard user.
The options are:
-**Enabled.** UIA programs, including Windows Remote Assistance, automatically disable the secure desktop for elevation prompts. If you do not disable the **User Account Control: Switch to the secure desktop when prompting for elevation** policy setting, the prompts appear on the interactive user's desktop instead of the secure desktop.
-**Disabled.** (Default) The secure desktop can be disabled only by the user of the interactive desktop or by disabling the **User Account Control: Switch to the secure desktop when prompting for elevation** policy setting.
UIA programs are designed to interact with Windows and application programs on behalf of a user. This policy setting allows UIA programs to bypass the secure desktop to increase usability in certain cases; however, allowing elevation requests to appear on the interactive desktop instead of the secure desktop can increase your security risk.
UIA programs must be digitally signed because they must be able to respond to prompts regarding security issues, such as the UAC elevation prompt. By default, UIA programs are run only from the following protected paths:
- ...\\Program Files, including subfolders
- ...\\Program Files (x86), including subfolders for 64-bit versions of Windows
- ...\\Windows\\System32
The **User Account Control: Only elevate UIAccess applications that are installed in secure locations** policy setting disables the requirement to be run from a protected path.
While this policy setting applies to any UIA program, it is primarily used in certain remote assistance scenarios, including the Windows Remote Assistance program in Windows 7.
If a user requests remote assistance from an administrator and the remote assistance session is established, any elevation prompts appear on the interactive user's secure desktop and the administrator's remote session is paused. To avoid pausing the remote administrator's session during elevation requests, the user may select the **Allow IT Expert to respond to User Account Control prompts** check box when setting up the remote assistance session. However, selecting this check box requires that the interactive user respond to an elevation prompt on the secure desktop. If the interactive user is a standard user, the user does not have the required credentials to allow elevation.
If you enable this policy setting, requests for elevation are automatically sent to the interactive desktop (not the secure desktop) and also appear on the remote administrator's view of the desktop during a remote assistance session. This allows the remote administrator to provide the appropriate credentials for elevation.
This policy setting does not change the behavior of the UAC elevation prompt for administrators.
If you plan to enable this policy setting, you should also review the effect of the **User Account Control: Behavior of the elevation prompt for standard users** policy setting. If it is configured as **Automatically deny elevation requests**, elevation requests are not presented to the user.
### User Account Control: Behavior of the elevation prompt for administrators in Admin Approval Mode
The **User Account Control: Behavior of the elevation prompt for administrators in Admin Approval Mode** policy setting controls the behavior of the elevation prompt for administrators.
The options are:
-**Elevate without prompting.** Allows privileged accounts to perform an operation that requires elevation without requiring consent or credentials.
**Note** Use this option only in the most constrained environments.
-**Prompt for credentials on the secure desktop.** When an operation requires elevation of privilege, the user is prompted on the secure desktop to enter a privileged user name and password. If the user enters valid credentials, the operation continues with the user's highest available privilege.
-**Prompt for consent on the secure desktop.** When an operation requires elevation of privilege, the user is prompted on the secure desktop to select either **Permit** or **Deny**. If the user selects **Permit**, the operation continues with the user's highest available privilege.
-**Prompt for credentials.** When an operation requires elevation of privilege, the user is prompted to enter an administrative user name and password. If the user enters valid credentials, the operation continues with the applicable privilege.
-**Prompt for consent.** When an operation requires elevation of privilege, the user is prompted to select either **Permit** or **Deny**. If the user selects **Permit**, the operation continues with the user's highest available privilege.
-**Prompt for consent for non-Windows binaries.** (Default) When an operation for a non-Microsoft application requires elevation of privilege, the user is prompted on the secure desktop to select either **Permit** or **Deny**. If the user selects **Permit**, the operation continues with the user's highest available privilege.
### User Account Control: Behavior of the elevation prompt for standard users
The **User Account Control: Behavior of the elevation prompt for standard users** policy setting controls the behavior of the elevation prompt for standard users.
The options are:
-**Automatically deny elevation requests.** When an operation requires elevation of privilege, a configurable access denied error message is displayed. An enterprise that is running desktops as standard user may choose this setting to reduce help desk calls.
-**Prompt for credentials on the secure desktop.** (Default) When an operation requires elevation of privilege, the user is prompted on the secure desktop to enter a different user name and password. If the user enters valid credentials, the operation continues with the applicable privilege.
-**Prompt for credentials.** When an operation requires elevation of privilege, the user is prompted to enter an administrative user name and password. If the user enters valid credentials, the operation continues with the applicable privilege.
### User Account Control: Detect application installations and prompt for elevation
The **User Account Control: Detect application installations and prompt for elevation** policy setting controls the behavior of application installation detection for the computer.
The options are:
-**Enabled.** (Default for home) When an application installation package is detected that requires elevation of privilege, the user is prompted to enter an administrative user name and password. If the user enters valid credentials, the operation continues with the applicable privilege.
-**Disabled.** (Default for enterprise) Application installation packages are not detected and prompted for elevation. Enterprises that are running standard user desktops and use delegated installation technologies such as Group Policy Software Installation or Systems Management Server (SMS) should disable this policy setting. In this case, installer detection is unnecessary.
### User Account Control: Only elevate executables that are signed and validated
The **User Account Control: Only elevate executables that are signed and validated** policy setting enforces public key infrastructure (PKI) signature checks for any interactive applications that request elevation of privilege. Enterprise administrators can control which applications are allowed to run by adding certificates to the Trusted Publishers certificate store on local computers.
The options are:
-**Enabled.** Enforces the PKI certification path validation for a given executable file before it is permitted to run.
-**Disabled.** (Default) Does not enforce PKI certification path validation before a given executable file is permitted to run.
### User Account Control: Only elevate UIAccess applications that are installed in secure locations
The **User Account Control: Only elevate UIAccess applications that are installed in secure locations** policy setting controls whether applications that request to run with a User Interface Accessibility (UIAccess) integrity level must reside in a secure location in the file system. Secure locations are limited to the following:
- ...\\Program Files, including subfolders
- ...\\Windows\\system32
- ...\\Program Files (x86), including subfolders for 64-bit versions of Windows
**Note** Windows enforces a PKI signature check on any interactive application that requests to run with a UIAccess integrity level regardless of the state of this security setting.
The options are:
-**Enabled.** (Default) If an application resides in a secure location in the file system, it runs only with UIAccess integrity.
-**Disabled.** An application runs with UIAccess integrity even if it does not reside in a secure location in the file system.
### User Account Control: Run all administrators in Admin Approval Mode
The **User Account Control: Run all administrators Admin Approval Mode** policy setting controls the behavior of all UAC policy settings for the computer. If you change this policy setting, you must restart your computer.
The options are:
-**Enabled.** (Default) Admin Approval Mode is enabled. This policy must be enabled and related UAC policy settings must also be set appropriately to allow the built-in Administrator account and all other users who are members of the **Administrators** group to run in Admin Approval Mode.
-**Disabled.** Admin Approval Mode and all related UAC policy settings are disabled.
**Note** If this policy setting is disabled, the Security Center notifies you that the overall security of the operating system has been reduced.
### User Account Control: Switch to the secure desktop when prompting for elevation
The **User Account Control: Switch to the secure desktop when prompting for elevation** policy setting controls whether the elevation request prompt is displayed on the interactive user's desktop or the secure desktop.
The options are:
-**Enabled.** (Default) All elevation requests go to the secure desktop regardless of prompt behavior policy settings for administrators and standard users.
-**Disabled.** All elevation requests go to the interactive user's desktop. Prompt behavior policy settings for administrators and standard users are used.
When this policy setting is enabled, it overrides the **User Account Control: Behavior of the elevation prompt for administrators in Admin Approval Mode** policy setting. The following table describes the behavior of the elevation prompt for each of the administrator policy settings when the **User Account Control: Switch to the secure desktop when prompting for elevation** policy setting is enabled or disabled.
| **Prompt for credentials on the secure desktop** | The prompt appears on the secure desktop. | The prompt appears on the secure desktop. |
| **Prompt for consent on the secure desktop** | The prompt appears on the secure desktop. | The prompt appears on the secure desktop. |
| **Prompt for credentials** | The prompt appears on the secure desktop. | The prompt appears on the interactive user's desktop. |
| **Prompt for consent** | The prompt appears on the secure desktop. | The prompt appears on the interactive user's desktop. |
| **Prompt for consent for non-Windows binaries** | The prompt appears on the secure desktop. | The prompt appears on the interactive user's desktop. |
When this policy setting is enabled, it overrides the **User Account Control: Behavior of the elevation prompt for standard users** policy setting. The following table describes the behavior of the elevation prompt for each of the standard user policy settings when the **User Account Control: Switch to the secure desktop when prompting for elevation** policy setting is enabled or disabled.
| Standard policy setting | Enabled | Disabled |
| - | - | - |
| **Automatically deny elevation requests** | No prompt. The request is automatically denied. | No prompt. The request is automatically denied. |
| **Prompt for credentials on the secure desktop** | The prompt appears on the secure desktop. | The prompt appears on the secure desktop. |
| **Prompt for credentials** | The prompt appears on the secure desktop. | The prompt appears on the interactive user's desktop. |
### User Account Control: Virtualize file and registry write failures to per-user locations
The **User Account Control: Virtualize file and registry write failures to per-user locations** policy setting controls whether application write failures are redirected to defined registry and file system locations. This policy setting mitigates applications that run as administrator and write run-time application data to %ProgramFiles%, %Windir%, %Windir%\\system32, or HKLM\\Software.
The options are:
-**Enabled.** (Default) Application write failures are redirected at run time to defined user locations for both the file system and registry.
-**Disabled.** Applications that write data to protected locations fail.
## Registry key settings
The registry keys are found in **HKEY\_LOCAL\_MACHINE\\SOFTWARE\\Microsoft\\Windows\\CurrentVersion\\Policies\\System**. For information about each of the registry keys, see the associated Group Policy description.
| EnableUIADesktopToggle | [User Account Control: Allow UIAccess applications to prompt for elevation without using the secure desktop](#user-account-control-allow-uiaccess-applications-to prompt-for-elevation-without-using-the-secure-desktop) | 0 (Default) = Disabled<br/>1 = Enabled |
| ConsentPromptBehaviorAdmin | [User Account Control: Behavior of the elevation prompt for administrators in Admin Approval Mode](#user-account-control-behavior-of-the-elevation-prompt-for-administrators-in-admin-approval-mode) | 0 = Elevate without prompting<br/>1 = Prompt for credentials on the secure desktop<br/>2 = Prompt for consent on the secure desktop<br/>3 = Prompt for credentials<br/>4 = Prompt for consent<br/>5 (Default) = Prompt for consent for non-Windows binaries<br/> |
| ConsentPromptBehaviorUser | [User Account Control: Behavior of the elevation prompt for standard users](#user-account-control-behavior-of-the-elevation-prompt-for-standard-users) | 0 = Automatically deny elevation requests<br/>1 = Prompt for credentials on the secure desktop<br/>3 (Default) = Prompt for credentials |
| EnableInstallerDetection | [User Account Control: Detect application installations and prompt for elevation](#user-account-control-detect-application-installations-and-prompt-for-elevation) | 1 = Enabled (default for home)<br/>0 = Disabled (default for enterprise) |
| ValidateAdminCodeSignatures | [User Account Control: Only elevate executables that are signed and validated](#user-account-control-only-elevate-executables-that-are-signed-and-validated) | 0 (Default) = Disabled<br/>1 = Enabled |
| EnableSecureUIAPaths | [User Account Control: Only elevate UIAccess applications that are installed in secure locations](#user-account-control-only-elevate-uiaccess-applications-that-are-installed-in-secure-locations) | 0 = Disabled<br/>1 (Default) = Enabled |
| EnableLUA | [User Account Control: Run all administrators in Admin Approval Mode](#user-account-control-run-all-administrators-in-admin-approval-mode) | 0 = Disabled<br/>1 (Default) = Enabled |
| PromptOnSecureDesktop | [User Account Control: Switch to the secure desktop when prompting for elevation](#user-account-control:-switch-to-the-secure-desktop-when-prompting-for-elevation) | 0 = Disabled<br/>1 (Default) = Enabled |
User Account Control (UAC) helps prevent malware from damaging a PC and helps organizations deploy a better-managed desktop. With UAC, apps and tasks always run in the security context of a non-administrator account, unless an administrator specifically authorizes administrator-level access to the system. UAC can block the automatic installation of unauthorized apps and prevent inadvertent changes to system settings.
## <a href="" id="bkmk-over"></a>
UAC allows all users to log on to their computers using a standard user account. Processes launched using a standard user token may perform tasks using access rights granted to a standard user. For instance, Windows Explorer automatically inherits standard user level permissions. Additionally, any apps that are started using Windows Explorer (for example, by double-clicking a shortcut) also run with the standard set of user permissions. Many apps, including those that are included with the operating system itself, are designed to work properly in this way.
Other apps, especially those that were not specifically designed with security settings in mind, often require additional permissions to run successfully. These types of apps are referred to as legacy apps. Additionally, actions such as installing new software and making configuration changes to the Windows Firewall, require more permissions than what is available to a standard user account.
When an app needs to run with more than standard user rights, UAC can restore additional user groups to the token. This enables the user to have explicit control of apps that are making system level changes to their computer or device.
Admin Approval Mode in UAC helps prevent malware from silently installing without an administrator's knowledge. It also helps protect from inadvertent system-wide changes. Lastly, it can be used to enforce a higher level of compliance where administrators must actively consent or provide credentials for each administrative process.
## <a href="" id="bkmk-new"></a>New and changed functionality
## New and changed functionality
To find out what's new in UAC for Windows10, see [User Account Control](../whats-new/user-account-control.md).
## In this section
<table>
<colgroup>
<colwidth="50%"/>
<colwidth="50%"/>
</colgroup>
<thead>
<trclass="header">
<thalign="left">Topic</th>
<thalign="left">Description</th>
</tr>
</thead>
<tbody>
<trclass="odd">
<tdalign="left"><p>[How User Account Control works](how-user-account-control-works.md)</p></td>
<tdalign="left"><p>User Account Control (UAC) is a fundamental component of Microsoft's overall security vision. UAC helps mitigate the impact of malware.</p></td>
</tr>
<trclass="even">
<tdalign="left"><p>[User Account Control security policy settings](user-account-control-security-policy-settings.md)</p></td>
<tdalign="left"><p>You can use security policies to configure how User Account Control works in your organization. They can be configured locally by using the Local Security Policy snap-in (secpol.msc) or configured for the domain, OU, or specific groups by Group Policy.</p></td>
</tr>
</tbody>
</table>
| Topic | Description |
| - | - |
| [How User Account Control works](how-user-account-control-works.md) | User Account Control (UAC) is a fundamental component of Microsoft's overall security vision. UAC helps mitigate the impact of malware. |
| [User Account Control security policy settings](user-account-control-security-policy-settings.md) | You can use security policies to configure how User Account Control works in your organization. They can be configured locally by using the Local Security Policy snap-in (secpol.msc) or configured for the domain, OU, or specific groups by Group Policy. |
| [User Account Control Group Policy and registry key settings](user-account-control-group-policy-and-registry-key-settings.md) | Here's a list of UAC Group Policy and registry key settings that your organization can use to manage UAC. |
@ -345,17 +345,16 @@ Table 3 lists specific malware threats and the mitigation that Windows10 provi
Table 3. Threats and Windows10 mitigations
<table>
<colgroup>
<colwidth="50%"/>
<colwidth="50%"/>
</colgroup>
<thead>
<trclass="header">
<thalign="left">Threat</th>
<thalign="left">Windows 10 mitigation</th>
</tr>
</thead>
<tbody>
<tbody><trclass="odd">
<tdalign="left"><p>"Man in the middle" attacks, when an attacker reroutes communications between two users through the attacker's computer without the knowledge of the two communicating users</p></td>
<tdalign="left"><p>Client connections to the Active Directory Domain Services default SYSVOL and NETLOGON shares on domain controllers now require SMB signing and mutual authentication (such as Kerberos).</p></td>
</tr>
<trclass="odd">
<tdalign="left"><p>Firmware bootkits replace the firmware with malware.</p></td>
<tdalign="left"><p>All certified PCs include a UEFI with Secure Boot, which requires signed firmware for updates to UEFI and Option ROMs.</p></td>
@ -395,6 +394,22 @@ Table 3. Threats and Windows10 mitigations
The sections that follow describe these improvements in more detail.
**SMB hardening improvements for SYSVOL and NETLOGON connections**
In Windows 10 and Windows Server 2016 Technical Preview, client connections to the Active Directory Domain Services default SYSVOL and NETLOGON shares on domain controllers now require Server Message Block (SMB) signing and mutual authentication (such as Kerberos).
- **What value does this change add?**
This change reduces the likelihood of man-in-the-middle attacks.
- **What works differently?**
If SMB signing and mutual authentication are unavailable, a Windows 10 or Windows Server 2016 computer won’t process domain-based Group Policy and scripts.
> **Note:** The registry values for these settings aren’t present by default, but the hardening rules still apply until overridden by Group Policy or other registry values.
For more information on these security improvements, (also referred to as UNC hardening), see [Microsoft Knowledge Base article 3000483](http://go.microsoft.com/fwlink/p/?LinkId=789216) and [MS15-011 & MS15-014: Hardening Group Policy](http://go.microsoft.com/fwlink/p/?LinkId=789215).
**Secure hardware**
Although Windows10 is designed to run on almost any hardware capable of running Windows8, Windows7, or WindowsVista, taking full advantage of Windows10 security requires advancements in hardware-based security, including UEFI with Secure Boot, CPU virtualization features (for example, Intel VT-x), CPU memory-protection features (for example, Intel VT-d), TPM, and biometric sensors.
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
jdeckerMS