Microsoft Office 365—Do you have a false sense of cloud security?

Through difficult times, some adversaries will find opportunities and COVID-19 has proven to be a ripe opportunity for them to target a new, expanding, remote workforce. While these threats morph and evolve, Microsoft’s Detection and Response Team (DART) finds ways to endure and help organizations become more resilient

Cloud environments are continuously being put to the test during this challenging period. DART has seen various security configurations in our customers’ cloud tenants. The one commonality:  administrators flip the switch on a few security tasks without genuinely understanding the process and procedures needed to ensure everything works as designed and consequently create gaps in defenses and opportunities for attackers to circumvent security controls. When it comes to defense-in-depth, these controls must work in concert with one another.

Three measures you should employ to improve the security of your cloud environment

This post describes three security measures you should employ for your Azure AD/Office 365 environment when first setting up a new tenant, or when tightening the reins on a well-established tenant.

  1. Create an emergency Global Administration account.
  2. Enable Multi-factor Authentication (MFA).
  3. Block legacy authentication.

1. Create an emergency Global Administrator account

An emergency Global Administrator account, also known as a “Break Glass Account”, is critical to the overall security posture of your tenant, and it prevents you from being accidentally locked out of your Azure Active Directory (Azure AD). Think about the consequences of your administrators getting locked out; you cannot sign in, activate users, assign licenses, or validate the actions happening in your tenant. Emergency access accounts are highly privileged and not assigned to specific users. These accounts must be excluded from your current security controls, and must have compensatory controls. These controls might include the following:

  • Only allowing the “Break Glass Account” to log in from a particular IP address range.
  • Implementing detection controls like enhanced alerting and/or monitoring the use of these accounts.

Use of emergency access accounts should be limited to true emergencies, when standard administrative accounts cannot be used. For detailed information, please see  https://docs.microsoft.com/en-us/azure/active-directory/users-groups-roles/directory-emergency-access.

2. Enable Multi-Factor Authentication (MFA)

Enabling MFA seems straightforward, right? Sadly, even today, it isn’t. You allow the Conditional Access Policy for the enablement of MFA, but for the sake of convenience,  permit exclusions to these policies, such as not enabling MFA for the Global Administrators or any of the other O365 workload (Exchange, SharePoint, OneDrive) Administrators and continue to enable Basic/Legacy Authentication. As a result, you now host an ineffective policy that puts your organization at tremendous risk. For detailed information, please see  https://docs.microsoft.com/en-us/azure/active-directory/authentication/tutorial-enable-azure-mfa.

Real-World Scenario—A large company enabled MFA for all global administrators. Unbeknownst to the rest of the team, a user modified the policy to exclude a global administrator account. This user’s act put the company at considerable risk; the account was eventually compromised using a trivial Password Spray attack. It is bad enough when a standard user with no elevated privileges is compromised— Global Administrator accounts have access to all of Azure AD and Office 365, so when this account was affected, the organization’s entire tenant was compromised. Monitoring and alerting for the implementation of persistence mechanisms, such as the creation of a new mailbox forwarding rule, would have also triggered a security alert and a full incident response investigation of the modified tenant. This incident also could have been easily avoided by merely monitoring and alerting for the creation of Global Administrator accounts and any changes to these accounts. The threat actor has leveraged all these techniques to essentially gain and maintain access to the organization’s tenant to achieve their mission objective for data exposure and exfiltration.

3. Block Legacy Authentication

Legacy authentication refers to protocols that use basic authentication, such as Exchange Web Services (EWS), POP, SMTP, IMAP, and MAPI. These protocols cannot enforce any type of second-factor authentication (e.g., MFA), which makes them a popular entry point for bad actors. As such, for MFA to be useful, you also need to block legacy authentication.

There are still risks once you’ve disabled legacy authentication and enabled MFA. From an operational standpoint, understanding the implications of disabling legacy authentication is critical. You could disrupt essential workflows and disrupt access to applications not written to support modern authentication (including dated Outlook clients).

So, what can you do? Identify which users and applications are currently using legacy authentication in your tenant via Azure AD Sign-in logs. Configure exclusions for applications that cannot be modified to support modern authentication. Also, ensure you configure the policies granularly for more robust security configurations, such as only allowing specific users and a particular IP range to use legacy authentication. This way, you can make access to legacy authentication more stringent where you must use it, and you can block legacy authentication in other scenarios. Configure your conditional access policy to be in a report-only mode to ensure you understand what will happen when you flip on the policy. For more information, please see https://docs.microsoft.com/en-us/azure/active-directory/conditional-access/block-legacy-authentication.

Bricks laid, next: the mortar

There is a multitude of adversary tactics and techniques for the infiltration of a cloud environment. Based on DART’s observations from the frontlines, implementing these three security controls will help ensure the front and back doors to your organization’s cloud environment remain locked. DART recommends assessing these vulnerability points regularly so that when a real threat strikes, your defense-in-depth approach of technical controls, detection-in-depth, and monitoring and alerts will prepare your staff to jump into action quickly.

In an upcoming blog post, we’ll dive into what we like to call the “Easy Button” approach to security defaults. These pre-configured security settings help defend your organization against frequent identity-related attacks, such as password spray, replay, and phishing, and provide additional mortar towards the security foundation of your cloud environment.

Want to learn more about DART (Detection and Response Team)? Read our past blog posts here.

To learn more about Microsoft Security solutions visit our website.  Bookmark the Security blog to keep up with our expert coverage on security matters. Also, follow us at @MSFTSecurity for the latest news and updates on cybersecurity.