Detecting Post-Compromise Threat Activity in Microsoft Cloud Environments

This Advisory uses the MITRE Adversarial Tactics, Techniques, and Common Knowledge (ATT&CK®) framework. See the ATT&CK for Enterprise for all referenced threat actor tactics and techniques.

This Alert is a companion alert to AA20-352A: Advanced Persistent Threat Compromise of Government Agencies, Critical Infrastructure, and Private Sector Organizations. AA20-352A primarily focuses on an advanced persistent threat (APT) actor’s compromise of SolarWinds Orion products as an initial access vector into networks of U.S. Government agencies, critical infrastructure entities, and private network organizations. As noted in AA20-352A, the Cybersecurity and Infrastructure Security Agency (CISA) has evidence of initial access vectors in addition to the compromised SolarWinds Orion products.

This Alert also addresses activity—irrespective of the initial access vector leveraged—that CISA attributes to an APT actor. Specifically, CISA has seen an APT actor using compromised applications in a victim’s Microsoft 365 (M365)/Azure environment. CISA has also seen this APT actor utilizing additional credentials and Application Programming Interface (API) access to cloud resources of private and public sector organizations. These tactics, techniques, and procedures (TTPs) feature three key components:

  • Compromising or bypassing federated identity solutions;
  • Using forged authentication tokens to move laterally to Microsoft cloud environments; and
  • Using privileged access to a victim’s cloud environment to establish difficult-to-detect persistence mechanisms for Application Programming Interface (API)-based access.

This Alert describes these TTPs and offers an overview of, and guidance on, available open-source tools—including a CISA-developed tool, Sparrow—for network defenders to analyze their Microsoft Azure Active Directory (AD), Office 365 (O365), and M365 environments to detect potentially malicious activity.

Note: this Alert describes artifacts—presented by these attacks—from which CISA has identified detectable evidence of the threat actor’s initial objectives. CISA continues to analyze the threat actor’s follow-on objectives.

Frequently, CISA has observed the APT actor gaining Initial Access [TA0001] to victims’ enterprise networks via compromised SolarWinds Orion products (e.g., Solorigate, Sunburst).[1] However, CISA is investigating instances in which the threat actor may have obtained initial access by Password Guessing [T1110.001], Password Spraying [T1110.003], and/or exploiting inappropriately secured administrative or service credentials (Unsecured Credentials [T1552]) instead of utilizing the compromised SolarWinds Orion products.

CISA observed this threat actor moving from user context to administrator rights for Privilege Escalation [TA0004] within a compromised network and using native Windows tools and techniques, such as Windows Management Instrumentation (WMI), to enumerate the Microsoft Active Directory Federated Services (ADFS) certificate-signing capability. This enumeration allows threat actors to forge authentication tokens (OAuth) to issue claims to service providers—without having those claims checked against the identity provider—and then to move laterally to Microsoft Cloud environments (Lateral Movement [TA0008]).

The threat actor has also used on-premises access to manipulate and bypass identity controls and multi-factor authentication. This activity demonstrates how sophisticated adversaries can use credentials from one portion of an organization to move laterally (Lateral Movement [TA0008]) through trust boundaries, evade defenses and detection (Defense Evasion [TA0005]), and steal sensitive data (Collection [TA0009]).

This level of compromise is challenging to remediate and requires a rigorous multi-disciplinary effort to regain administrative control before recovering.


Guidance on identifying affected SolarWinds software is well documented.[2] However—once an organization identifies a compromise via SolarWinds Orion products or other threat actor TTPs—identifying follow-on activity for on-premises networks requires fine-tuned network and host-based forensics.

The nature of cloud forensics is unique due to the growing and rapidly evolving technology footprints of major vendors. Microsoft’s O365 and M365 environments have built-in capabilities for detecting unusual activity. Microsoft also provides premium services (Advanced Threat Protection [ATP] and Azure Sentinel), which enable network defenders to investigate TTPs specific to the Solorigate activity.[3]

Detection Tools

CISA is providing examples of detection tools for informational purposes only. CISA does not endorse any commercial product or service, including any subjects of analysis. Any reference to specific commercial products, processes, or services does not constitute or imply their endorsement, recommendation, or favoring by CISA.

There are a number of open-source tools available to investigate adversary activity in Microsoft cloud environments and to detect unusual activity, service principals, and application activity.[4] Publicly available PowerShell tools that network defenders can use to investigate M365 and Microsoft Azure include:

  • CISA’s Sparrow,
  • Open-source utility Hawk, and
  • CrowdStrike’s Azure Reporting Tool (CRT).

Additionally, Microsoft’s Office 365 Management API and Graph API provide an open interface for ingesting telemetry and evaluating service configurations for signs of anomalous activity and intrusion.

Note: these open-source tools are highlighted and explained to assist with on-site investigation and remediation in cloud environments but are not all-encompassing. Open source tools can be complemented by services such as Azure Sentinel, a Microsoft premium service that provides comprehensive analysis tools, including custom detections for the activity indicated.

General Guidance on Using Detection Tools

  1. Audit the creation and use of service principal credentials. Look for unusual application usage, such as use of dormant applications.
  2. Audit the assignment of credentials to applications that allow non-interactive sign-in by the application. Look for unexpected trust relationships added to the Azure Active Directory.
  3. Download the interactive sign-ins from the Azure admin portal or use the Microsoft Sentinel product. Review new token validation time periods with high values and investigate whether it was a legitimate change or an attempt to gain persistence by a threat actor.


CISA created Sparrow to help network defenders detect possible compromised accounts and applications in the Azure/M365 environment. The tool focuses on the narrow scope of user and application activity endemic to identity- and authentication-based attacks seen recently in multiple sectors. It is neither comprehensive nor exhaustive of available data. It is intended to narrow a larger set of available investigation modules and telemetry to those specific to recent attacks on federated identity sources and applications.

CISA advises Sparrow users to take the following actions.

  1. Use Sparrow to detect any recent domain authentication or federation modifications.
    1. Domain and federation modification operations are uncommon and should be investigated.
  2. Examine logs for new and modified credentials applied to applications and service principals; delineate for the credential type. Sparrow can be used to detect the modification of service principals and application credentials.
    1. Create a timeline for all credential changes, focusing on recent wholesale changes.
    2. Review the “top actors” for activity in the environment and the number of credential modifications performed.
    3. Monitor changes in application and service principal credentials.
    4. Investigate any instances of excessive permissions being granted, including, but not limited to, Exchange Online, Microsoft Graph, and Azure AD Graph.
  3. Use Sparrow to detect privilege escalation, such as adding a service principal, user, or group to a privileged role.
  4. Use Sparrow to detect OAuth consent and users’ consent to applications, which is useful for interpreting changes in adversary TTPs.
  5. Use Sparrow to identify anomalous Security Assertion Markup Language (SAML) token sign-ins by pivoting on the unified audit log UserAuthenticationValue of 16457, which is an indicator of how a SAML token was built and is a potential indicator for forged SAML tokens.
    1. Note that this TTP has not been the subject of significant published security research but may indicate an unusual usage of a token, such as guest access for external partners to M365 resources.
  6. Review the PowerShell logs that Sparrow exports.
    1. Review PowerShell mailbox sign-ins and validate that the logins are legitimate actions.
    2. Review PowerShell usage for users with PowerShell in the environment.
  7. Use Sparrow to check the Graph API application permissions of all service principals and applications in M365/Azure AD.
    1. Investigate unusual activity regarding Microsoft Graph API permissions (using either the legacy or Graph is used frequently as part of these TTPs, often to access and manipulate mailbox resources.
  8. Review Sparrow’s listed tenant’s Azure AD domains, to see if the domains have been modified.
  9. For customers with G5 or E5 licensing levels, review MailItemsAccessed for insight into what application identification (ID) was used for accessing users’ mailboxes. Use Sparrow to query for a specific application ID using the app id investigation capability, which will check to see if it is accessing mail or file items.
    1. The MailItemsAccessed event provides audibility for mailbox data accessed via mail protocols or clients.
    2. By analyzing the MailItemsAccessed action, incident responders can determine which user mailbox items have been accessed and potentially exfiltrated by a threat actor. This event will be recorded even in some situations where the message was not necessarily read interactively (e.g., bind or sync).[5]
    3. The resulting suspicious application ID can provide incident responders with a pivot to detect other suspicious applications that require additional analysis.
    4. Check for changes to applications with regards to the accessing of resources such as mail or file items.


Hawk is an open-source, PowerShell-driven, community-developed tool network defenders can use to quickly and easily gather data from O365 and Azure for security investigations. Incident responders and network defenders can investigate specific user principals or the entire tenant. Data it provides include IP addresses and sign-in data. Additionally, Hawk can track IP usage for concurrent login situations.

Hawk users should review login details for administrator accounts and take the following steps.

  1.  Investigate high-value administrative accounts to detect anomalous or unusual activity (Global Admins).
  2. Enable PowerShell logging, and evaluate PowerShell activity in the environment not used for traditional or expected purposes.
    1. PowerShell logging does not reveal the exact cmdlet that was run on the tenant.
  3. Look for users with unusual sign-in locations, dates, and times.
  4. Check permissions of service principals and applications in M365/Azure AD.
  5. Detect the frequency of resource access from unusual places. Use the tool to pivot to a trusted application and see if it is accessing mail or file items.
  6. Review mailbox rules and recent mailbox rule changes.

CrowdStrike Azure Reporting Tool

CrowdStrike’s Azure Reporting Tool (CRT) can help network defenders analyze their Microsoft Azure AD and M365 environment to help organizations analyze permissions in their Azure AD tenant and service configuration. This tool has minor overlap with Sparrow; it shows unique items, but it does not cover the same areas. CISA is highlighting this tool because it is one of the only free, open-source tools available to investigate this activity and could be used to complement Sparrow.

Detection Tool Distinctions

  • Sparrow differs from CRT by looking for specific indicators of compromise associated with the recent attacks.
  • CRT focuses on the tenant’s Azure AD permissions and Exchange Online configuration settings instead of the unified audit log, which gives it a different output from Sparrow or Hawk.
  • CRT returns the same broad scope of application/delegated permissions for service principals and applications as Hawk.
  • As part of its investigation, Sparrow homes in on a narrow set of application permissions given to the Graph API, which is common to the recent attacks.
  • CRT looks at Exchange Online federation configuration and federation trust, while Sparrow focuses on listing Azure AD domains.
  • Among the items network defenders can use CRT to review are delegated permissions and application permissions, federation configurations, federation trusts, mail forwarding rules, service principals, and objects with KeyCredentials.

Detection Methods

Microsoft breaks the threat actor’s recent activity into four primary stages, which are described below along with associated detection methods. Microsoft describes these stages as beginning with all activity after the compromise of the on-premises identity solution, such as ADFS.[6]

Note: this step provides an entry vector to cloud technology environments, and is unnecessary when the threat actor has compromised an identity solution or credential that allows the APT direct access to the cloud(e.g., without leveraging the SolarWinds Orion vulnerability).

Stage 1: Forging a trusted authentication token used to access resources that trust the on-premises identity provider

These attacks (often referred to as “Golden Security Assertion Markup Language” attacks) can be analyzed using a combination of cloud-based and standard on-premises techniques.[7] For example, network defenders can use OAuth claims for specific principals made at the Azure AD level and compare them to the on-premises identity.

Export sign-in logs from the Azure AD portal and look at the Authentication Method field.

Note: at, click on a user and review the authentication details (e.g., date, method, result). Without Sentinel, this is the only way to get these logs, which are critical for this effort.

Detection Method 1: Correlating service provider login events with corresponding authentication events in Active Directory Federation Services (ADFS) and Domain Controllers

Using SAML single sign-on, search for any logins to service providers that do not have corresponding event IDs 4769, 1200, and 1202 in the domain.

Detection Method 2: Identifying certificate export events in ADFS

Look for:

  1. The IP address and Activity_ID in EventCode 410 and the Activity_ID and Instance_ID in EventCode 500.
  2. Export-PfxCertificate or certutil-exportPFX in Event IDs 4103 and 4104, which may include detection of a certificate extraction technique.
  3. Deleted certificate extraction with ADFSdump performed using Sysmon Event ID 18 with the pipe name microsoft##widtsqlquery (exclude processes regularly making this pipe connection on the machine).
  4. Event ID 307 (The Federation Service configuration was changed), which can be correlated to relevant Event ID 510 with the same instance ID for change details (Event ID 510 with the same Instance ID could be more than one event per single Event ID 307 event).

Detection Method 3: Customizing SAML response to identify irregular access

This method serves as prevention for the future (and would only detect future, not past, activity), as it helps identify irregularities from the point of the change forward. Organizations can modify SAML responses to include custom elements for each service provider to monitor and detect any anomalous requests.[8]

Detection Method 4: Detecting malicious ADFS trust modification

A threat actor who gains administrative access to ADFS can add a new, trusted ADFS rather than extracting the certificate and private key as part of a standard Golden SAML attack.[9]
Network defenders should look for:

  1. Event ID 307 (The Federation Service configuration was changed), which can be correlated to relevant Event ID 510 with the same Instance ID for change details. (Event ID 510 with the same Instance ID could be more than one event per single Event ID 307 event.)
    1. Review events, particularly searching for Configuration: Type: IssuanceAuthority where Property Value references an unfamiliar domain.
  2. Possible activity of an interrogating ADFS host by using ADFS PowerShell plugins. Look for changes in the federation trust environment that would indicate new ADFS sources.

Stage 2: Using the forged authentication token to create configuration changes in the Service Provider, such as Azure AD (establishing a foothold)

After the threat actor has compromised the on-premises identity provider, they identify their next series of objectives by reviewing activity in the Microsoft Cloud activity space (Microsoft Azure and M365 tenants).

The threat actor uses the ability to forge authentication tokens to establish a presence in the cloud environment. The actor adds additional credentials to an existing service principal. Once the threat actor has impersonated a privileged Azure AD account, they are likely to further manipulate the Azure/M365 environment (action on objectives in the cloud).

Network defenders should take the following steps.

  1. Audit the creation and use of service principal and application credentials. Sparrow will detect modifications to these credentials.
    1. Look for unusual application usage, such as dormant or forgotten applications being used again.
    2. Audit the assignment of credentials to applications that allow non-interactive sign-in by the application.
  2. Look for unexpected trust relationships that have been added to Azure AD. (Download the last 30 days of non-interactive sign-ins from the Azure portal or use Azure Sentinel.).[10]
  3. Use Hawk (and any sub-modules available) to run an investigation on a specific user. Hawk will provide IP addresses, sign-in data, and other data. Hawk can also track IP usage in concurrent login situations.
  4. Review login details for administrator accounts (e.g., high-value administrative accounts, such as Global Admins). Look for unusual sign-in locations, dates, and times.
  5. Review new token validation time periods with high values and investigate whether the changes are legitimate or a threat actor’s attempts to gain persistence.

Stage 3: Acquiring an OAuth access token for the application using the forged credentials added to an existing application or service principal and calling APIs with the permissions assigned to that application

In some cases, the threat actor has been observed adding permissions to existing applications or service principals. Additionally the actor has been seen establishing new applications or service principals briefly and using them to add permissions to the existing applications or service principals, possibly to add a layer of indirection (e.g., using it to add a credential to another service principal, and then deleting it).[11]

Network defenders should use Sparrow to:

  1. Examine highly privileged accounts; specifically using sign-in logs, look for unusual sign-in locations, dates, and times.
  2. Create a timeline for all credential changes.
  3. Monitor changes in application credentials (the script will export into csv named AppUpdate_Operations_Export).
  4. Detect service principal credentials change and service principal change (e.g., if an actor adds new permissions or expands existing permissions).
    1. Export and view this activity via the ServicePrincipal_Operations_Export.
  5. Record OAuth consent and consent to applications
    1. Export and view this record via the Consent_Operations_Export file.
  6. Investigate instances of excessive high permissions, including, but not limited to Exchange Online, Microsoft Graph, and Azure AD Graph.
    1. Review Microsoft Graph API permissions granted to service principals.
    2. Export and view this activity via the ApplicationGraphPermissions csv file.
      1. Note: Hawk can also return the full list of service principal permissions for further investigation.
    3. Review top actors and the amount of credential modifications performed.
    4. Monitor changes in application credentials.
  7. Identify manipulation of custom or third-party applications.
    1. Network defenders should review the catalog of custom or third-party vendors with applications in the Microsoft tenant and perform the above interrogation principles on those applications and trusts.
  8. Review modifications to federation trust settings.
    1. Review new token validation time periods with high values and investigate whether this was a legitimate change or an attempt to gain persistence by the threat actor.
      1. The script detects the escalation of privileges, including the addition of Service Principals (SP) to privileged roles. Export this data into csv called AppRoleAssignment_Operations_Export.

Stage 4: Once access has been established, the threat actor Uses Microsoft Graph API to conduct action on objectives from an external RESTful API (queries impersonating existing applications).

Network defenders should:

  1. In MailItemsAccessed operations, found within the Unified Audit Log (UAL), review the application ID used (requires G5 or E5 license for this specific detail).
  2. Query the specific application ID, using the Sparrow script’s app ID investigation capability to interrogate mail and file items accessed for that applicationID (Use the application ID utility for any other suspicious apps that require additional analysis.).
  3. Check the permissions of an application in M365/Azure AD using Sparrow.
    1. Hawk will return Azure_Application_Audit, and Sparrow will return ApplicationGraphPermissions.
    2. Network defenders will see the IP address that Graph API uses.
    3. Note: the Microsoft IP address may not show up as a virtual private server/anonymized endpoint.
  4. Investigate a specific service principal, if it is a user-specific user account, in Hawk. This activity is challenging to see without Azure Sentinel or manually downloading and reviewing logs from the sign-in portal.

Microsoft Telemetry Nuances

The existing tools and techniques used to evaluate cloud-based telemetry sources present challenges not represented in traditional forensic techniques. Primarily, the amount of telemetry retention is far less than the traditional logging facilities of on-premises data sources. Threat actor activity that is more than 90 days old is unlikely to have been saved by traditional sources or be visible with the Microsoft M365 Management API or in the UAL.

Service principal logging is available using the Azure Portal via the “Service Principal Sign-ins” feature. Enable settings in the Azure Portal (see “Diagnostic Setting”) to ingest logs into Sentinel or a third-party security information and event management (SIEM) tool. An Azure Premium P1 or Premium P2 license is necessary to access this setting as well as other features, such as a log analytics workspace, storage account, or event hub.[12] These logs must be downloaded manually if not ingested by one of the methods listed in the Detection Methods section.

Global Administrator rights are often required by tools other than Hawk and Sparrow to evaluate M365 cloud security posture. Logging capability and visibility of data varies by licensing models and subscription to premium services, such as Microsoft Defender for O365 and Azure Sentinel. According to CrowdStrike, “There was an inability to audit via API, and there is the requirement for global admin rights to view important information which we found to be excessive. Key information should be easily accessible.”[13]

Documentation for specific event codes, such as UserAuthenticationMethod 16457, which may indicate a suspicious SAML token forgery, is no longer available in the M365 Unified Access Log. Auditing narratives on some events no longer exist as part of core Microsoft documentation sources.

The use of industry-standard SIEMs for log detection is crucial for providing historical context for threat hunting in Microsoft cloud environments. Standard G3/E3 licenses only provide 90 days of auditing; with the advanced auditing license that is provided with a G5/E5 license, audit logs can be extended to retain information for a year. CISA notes that this license change is proactive, rather than reactive: it allows enhanced visibility and features for telemetry from the moment of integration but does not provide retroactive visibility on previous events or historical context.

A properly configured SIEM can provide:

  1. Longer term storage of log data.
  2. Cross correlation of log data with endpoint data and network data (such as those produced by ADFS servers), endpoint detection and response data, and identity provider information.
  3. Ability to query use of application connectors in Azure.

Built-in tools, such as Microsoft Cloud Services and M365 applications, provide much of the same visibility available from custom tools and are mapped to the MITRE ATT&CK framework and easy-to-understand dashboards.[14] However, these tools often do not have the ability to pull historical data older than seven days. Therefore, storage solutions that appropriately meet governance standards and usability metrics for analysts for the SIEM must be carefully planned and arranged.


Advanced Persistent Threat Compromise of Government Agencies, Critical Infrastructure, and Private Sector Organizations

This Alert uses the MITRE Adversarial Tactics, Techniques, and Common Knowledge (ATT&CK®) version 8 framework. See the ATT&CK for Enterprise version 8 for all referenced threat actor tactics and techniques.

The Cybersecurity and Infrastructure Security Agency (CISA) is aware of compromises of U.S. government agencies, critical infrastructure entities, and private sector organizations by an advanced persistent threat (APT) actor beginning in at least March 2020. This APT actor has demonstrated patience, operational security, and complex tradecraft in these intrusions. CISA expects that removing this threat actor from compromised environments will be highly complex and challenging for organizations.

One of the initial access vectors for this activity is a supply chain compromise of the following SolarWinds Orion products (see Appendix A).

  • Orion Platform 2019.4 HF5, version 2019.4.5200.9083
  • Orion Platform 2020.2 RC1, version 2020.2.100.12219
  • Orion Platform 2020.2 RC2, version 2020.2.5200.12394
  • Orion Platform 2020.2, 2020.2 HF1, version 2020.2.5300.12432

Note: CISA has evidence of additional initial access vectors, other than the SolarWinds Orion platform; however, these are still being investigated. CISA will update this Alert as new information becomes available.

On December 13, 2020, CISA released Emergency Directive 21-01: Mitigate SolarWinds Orion Code Compromise, ordering federal civilian executive branch departments and agencies to disconnect affected devices. Note: this Activity Alert does not supersede the requirements of Emergency Directive 21-01 (ED-21-01) and does not represent formal guidance to federal agencies under ED 21-01.

CISA has determined that this threat poses a grave risk to the Federal Government and state, local, tribal, and territorial governments as well as critical infrastructure entities and other private sector organizations. CISA advises stakeholders to read this Alert and review the enclosed indicators (see Appendix B).

Key Takeaways

  • This is a patient, well-resourced, and focused adversary that has sustained long duration activity on victim networks.
  • The SolarWinds Orion supply chain compromise is not the only initial infection vector this APT actor leveraged.
  • Not all organizations that have the backdoor delivered through SolarWinds Orion have been targeted by the adversary with follow-on actions.
  • Organizations with suspected compromises need to be highly conscious of operational security, including when engaging in incident response activities and planning and implementing remediation plans. 

Click here for a PDF version of this report.


CISA is aware of compromises, which began at least as early as March 2020, at U.S. government agencies, critical infrastructure entities, and private sector organizations by an APT actor. This threat actor has demonstrated sophistication and complex tradecraft in these intrusions. CISA expects that removing the threat actor from compromised environments will be highly complex and challenging. This adversary has demonstrated an ability to exploit software supply chains and shown significant knowledge of Windows networks. It is likely that the adversary has additional initial access vectors and tactics, techniques, and procedures (TTPs) that have not yet been discovered. CISA will continue to update this Alert and the corresponding indicators of compromise (IOCs) as new information becomes available.

Initial Infection Vectors [TA0001]

CISA is investigating incidents that exhibit adversary TTPs consistent with this activity, including some where victims either do not leverage SolarWinds Orion or where SolarWinds Orion was present but where there was no SolarWinds exploitation activity observed. Volexity has also reported publicly that they observed the APT using a secret key that the APT previously stole in order to generate a cookie to bypass the Duo multi-factor authentication protecting access to Outlook Web App (OWA).[1] Volexity attributes this intrusion to the same activity as the SolarWinds Orion supply chain compromise, and the TTPs are consistent between the two. This observation indicates that there are other initial access vectors beyond SolarWinds Orion, and there may still be others that are not yet known.

SolarWinds Orion Supply Chain Compromise

SolarWinds Orion is an enterprise network management software suite that includes performance and application monitoring and network configuration management along with several different types of analyzing tools. SolarWinds Orion is used to monitor and manage on-premise and hosted infrastructures. To provide SolarWinds Orion with the necessary visibility into this diverse set of technologies, it is common for network administrators to configure SolarWinds Orion with pervasive privileges, making it a valuable target for adversary activity.

The threat actor has been observed leveraging a software supply chain compromise of SolarWinds Orion products[2] (see Appendix A). The adversary added a malicious version of the binary solarwinds.orion.core.businesslayer.dll into the SolarWinds software lifecycle, which was then signed by the legitimate SolarWinds code signing certificate. This binary, once installed, calls out to a victim-specific avsvmcloud[.]com domain using a protocol designed to mimic legitimate SolarWinds protocol traffic. After the initial check-in, the adversary can use the Domain Name System (DNS) response to selectively send back new domains or IP addresses for interactive command and control (C2) traffic. Consequently, entities that observe traffic from their SolarWinds Orion devices to avsvmcloud[.]com should not immediately conclude that the adversary leveraged the SolarWinds Orion backdoor. Instead, additional investigation is needed into whether the SolarWinds Orion device engaged in further unexplained communications. If additional Canonical Name record (CNAME) resolutions associated with the avsvmcloud[.]com domain are observed, possible additional adversary action leveraging the back door has occurred.

Based on coordinated actions by multiple private sector partners, as of December 15, 2020, avsvmcloud[.]com resolves to 20.140.0[.]1, which is an IP address on the Microsoft blocklist. This negates any future use of the implants and would have caused communications with this domain to cease. In the case of infections where the attacker has already moved C2 past the initial beacon, infection will likely continue notwithstanding this action.

SolarWinds Orion typically leverages a significant number of highly privileged accounts and access to perform normal business functions. Successful compromise of one of these systems can therefore enable further action and privileges in any environment where these accounts are trusted.

Anti-Forensic Techniques

The adversary is making extensive use of obfuscation to hide their C2 communications. The adversary is using virtual private servers (VPSs), often with IP addresses in the home country of the victim, for most communications to hide their activity among legitimate user traffic. The attackers also frequently rotate their “last mile” IP addresses to different endpoints to obscure their activity and avoid detection.

FireEye has reported that the adversary is using steganography (Obfuscated Files or Information: Steganography [T1027.003]) to obscure C2 communications.[3] This technique negates many common defensive capabilities in detecting the activity. Note: CISA has not yet been able to independently confirm the adversary’s use of this technique.

According to FireEye, the malware also checks for a list of hard-coded IPv4 and IPv6 addresses—including RFC-reserved IPv4 and IPv6 IP—in an attempt to detect if the malware is executed in an analysis environment (e.g., a malware analysis sandbox); if so, the malware will stop further execution. Additionally, FireEye analysis identified that the backdoor implemented time threshold checks to ensure that there are unpredictable delays between C2 communication attempts, further frustrating traditional network-based analysis.

While not a full anti-forensic technique, the adversary is heavily leveraging compromised or spoofed tokens for accounts for lateral movement. This will frustrate commonly used detection techniques in many environments. Since valid, but unauthorized, security tokens and accounts are utilized, detecting this activity will require the maturity to identify actions that are outside of a user’s normal duties. For example, it is unlikely that an account associated with the HR department would need to access the cyber threat intelligence database.

Taken together, these observed techniques indicate an adversary who is skilled, stealthy with operational security, and is willing to expend significant resources to maintain covert presence.

Privilege Escalation and Persistence [TA0004, TA0003]

The adversary has been observed using multiple persistence mechanisms across a variety of intrusions. CISA has observed the threat actor adding authentication tokens and credentials to highly privileged Active Directory domain accounts as a persistence and escalation mechanism. In many instances, the tokens enable access to both on-premise and hosted resources. Microsoft has released a query that can help detect this activity.[4]

Microsoft reported that the actor has added new federation trusts to existing infrastructure, a technique that CISA believes was utilized by a threat actor in an incident to which CISA has responded. Where this technique is used, it is possible that authentication can occur outside of an organization’s known infrastructure and may not be visible to the legitimate system owner. Microsoft has released a query to help identify this activity.[5]

User Impersonation

The adversary’s initial objectives, as understood today, appear to be to collect information from victim environments. One of the principal ways the adversary is accomplishing this objective is by compromising the Security Assertion Markup Language (SAML) signing certificate using their escalated Active Directory privileges. Once this is accomplished, the adversary creates unauthorized but valid tokens and presents them to services that trust SAML tokens from the environment. These tokens can then be used to access resources in hosted environments, such as email, for data exfiltration via authorized application programming interfaces (APIs).

CISA has observed in its incident response work adversaries targeting email accounts belonging to key personnel, including IT and incident response personnel.

These are some key functions and systems that commonly use SAML.

  • Hosted email services
  • Hosted business intelligence applications
  • Travel systems
  • Timecard systems
  • File storage services (such as SharePoint)

Detection: Impossible Logins

The adversary is using a complex network of IP addresses to obscure their activity, which can result in a detection opportunity referred to as “impossible travel.” Impossible travel occurs when a user logs in from multiple IP addresses that are a significant geographic distance apart (i.e., a person could not realistically travel between the geographic locations of the two IP addresses during the time period between the logins). Note: implementing this detection opportunity can result in false positives if legitimate users apply virtual private network (VPN) solutions before connecting into networks.

Detection: Impossible Tokens

The following conditions may indicate adversary activity.

  • Most organizations have SAML tokens with 1-hour validity periods. Long SAML token validity durations, such as 24 hours, could be unusual.
  • The SAML token contains different timestamps, including the time it was issued and the last time it was used. A token having the same timestamp for when it was issued and when it was used is not indicative of normal user behavior as users tend to use the token within a few seconds but not at the exact same time of issuance.
  • A token that does not have an associated login with its user account within an hour of the token being generated also warrants investigation.

Operational Security

Due to the nature of this pattern of adversary activity—and the targeting of key personnel, incident response staff, and IT email accounts—discussion of findings and mitigations should be considered very sensitive, and should be protected by operational security measures. An operational security plan needs to be developed and socialized, via out-of-band communications, to ensure all staff are aware of the applicable handling caveats.

Operational security plans should include:

  • Out-of-band communications guidance for staff and leadership;
  • An outline of what “normal business” is acceptable to be conducted on the suspect network;
  • A call tree for critical contacts and decision making; and
  • Considerations for external communications to stakeholders and media.

MITRE ATT&CK® Techniques

CISA assesses that the threat actor engaged in the activities described in this Alert uses the below-listed ATT&CK techniques.

  • Query Registry [T1012]
  • Obfuscated Files or Information [T1027]
  • Obfuscated Files or Information: Steganography [T1027.003]
  • Process Discovery [T1057]
  • Indicator Removal on Host: File Deletion [T1070.004]
  • Application Layer Protocol: Web Protocols [T1071.001]
  • Application Layer Protocol: DNS [T1071.004]
  • File and Directory Discovery [T1083]
  • Ingress Tool Transfer [T1105]
  • Data Encoding: Standard Encoding [T1132.001]
  • Supply Chain Compromise: Compromise Software Dependencies and Development Tools [T1195.001]
  • Supply Chain Compromise: Compromise Software Supply Chain [T1195.002]
  • Software Discovery [T1518]
  • Software Discovery: Security Software [T1518.001]
  • Create or Modify System Process: Windows Service [T1543.003]
  • Subvert Trust Controls: Code Signing [T1553.002]
  • Dynamic Resolution: Domain Generation Algorithms [T1568.002]
  • System Services: Service Execution [T1569.002]
  • Compromise Infrastructure [T1584]

SolarWinds Orion Owners

Owners of vulnerable SolarWinds Orion products will generally fall into one of three categories.

  • Category 1 includes those who do not have the identified malicious binary. These owners can patch their systems and resume use as determined by and consistent with their internal risk evaluations.
  • Category 2 includes those who have identified the presence of the malicious binary—with or without beaconing to avsvmcloud[.]com. Owners with malicious binary whose vulnerable appliances only unexplained external communications are with avsvmcloud[.]com—a fact that can be verified by comprehensive network monitoring for the device—can harden the device, re-install the updated software from a verified software supply chain, and resume use as determined by and consistent with a thorough risk evaluation.
  • Category 3 includes those with the binary beaconing to avsvmcloud[.]com and secondary C2 activity to a separate domain or IP address. If you observed communications with avsvmcloud[.]com that appear to suddenly cease prior to December 14, 2020— not due to an action taken by your network defenders—you fall into this category. Assume the environment has been compromised, and initiate incident response procedures immediately.

Compromise Mitigations

If the adversary has compromised administrative level credentials in an environment—or if organizations identify SAML abuse in the environment, simply mitigating individual issues, systems, servers, or specific user accounts will likely not lead to the adversary’s removal from the network. In such cases, organizations should consider the entire identity trust store as compromised. In the event of a total identity compromise, a full reconstitution of identity and trust services is required to successfully remediate. In this reconstitution, it bears repeating that this threat actor is among the most capable, and in many cases, a full rebuild of the environment is the safest action.

SolarWinds Orion Specific Mitigations

The following mitigations apply to networks using the SolarWinds Orion product. This includes any information system that is used by an entity or operated on its behalf.

Organizations that have the expertise to take the actions in Step 1 immediately should do so before proceeding to Step 2. Organizations without this capability should proceed to Step 2. Federal civilian executive branch agencies should ignore the below and refer instead to Emergency Directive 21-01 (and forthcoming associated guidance) for mitigation steps.

  • Step 1
    • Forensically image system memory and/or host operating systems hosting all instances of affected versions of SolarWinds Orion. Analyze for new user or service accounts, privileged or otherwise.
    • Analyze stored network traffic for indications of compromise, including new external DNS domains to which a small number of agency hosts (e.g., SolarWinds systems) have had connections.
  • Step 2
    • Affected organizations should immediately disconnect or power down affected all instances of affected versions of SolarWinds Orion from their network.
    • Additionally:
      • Block all traffic to and from hosts, external to the enterprise, where any version of SolarWinds Orion software has been installed.
      • Identify and remove all threat actor-controlled accounts and identified persistence mechanisms.  
  • Step 3  
    • Only after all known threat actor-controlled accounts and persistence mechanisms have been removed:

See Joint Alert on Technical Approaches to Uncovering and Remediating Malicious Activity for more information on incident investigation and mitigation steps based on best practices.

CISA will update this Alert, as information becomes available and will continue to provide technical assistance, upon request, to affected entities as they work to identify and mitigate potential compromises.


Cyber Actors Target K-12 Distance Learning Education to Cause Disruptions and Steal Data

This Joint Cybersecurity Advisory was coauthored by the Federal Bureau of Investigation (FBI), the Cybersecurity and Infrastructure Security Agency (CISA), and the Multi-State Information Sharing and Analysis Center (MS-ISAC).

The FBI, CISA, and MS-ISAC assess malicious cyber actors are targeting kindergarten through twelfth grade (K-12) educational institutions, leading to ransomware attacks, the theft of data, and the disruption of distance learning services. Cyber actors likely view schools as targets of opportunity, and these types of attacks are expected to continue through the 2020/2021 academic year. These issues will be particularly challenging for K-12 schools that face resource limitations; therefore, educational leadership, information technology personnel, and security personnel will need to balance this risk when determining their cybersecurity investments.

Click here for a PDF version of this report.

As of December 2020, the FBI, CISA, and MS-ISAC continue to receive reports from K-12 educational institutions about the disruption of distance learning efforts by cyber actors.


The FBI, CISA, and MS-ISAC have received numerous reports of ransomware attacks against K-12 educational institutions. In these attacks, malicious cyber actors target school computer systems, slowing access, and—in some instances—rendering the systems inaccessible for basic functions, including distance learning. Adopting tactics previously leveraged against business and industry, ransomware actors have also stolen—and threatened to leak—confidential student data to the public unless institutions pay a ransom.

According to MS-ISAC data, the percentage of reported ransomware incidents against K-12 schools increased at the beginning of the 2020 school year. In August and September, 57% of ransomware incidents reported to the MS-ISAC involved K-12 schools, compared to 28% of all reported ransomware incidents from January through July.

The five most common ransomware variants identified in incidents targeting K-12 schools between January and September 2020—based on open source information as well as victim and third-party incident reports made to MS-ISAC—are Ryuk, Maze, Nefilim, AKO, and Sodinokibi/REvil.


Figure 1 identifies the top 10 malware strains that have affected state, local, tribal, and territorial (SLTT) educational institutions over the past year (up to and including September 2020). Note: These malware variants are purely opportunistic as they not only affect educational institutions but other organizations as well.

ZeuS and Shlayer are among the most prevalent malware affecting K-12 schools.

  • ZeuS is a Trojan with several variants that targets Microsoft Windows operating systems. Cyber actors use ZeuS to infect target machines and send stolen information to command-and-control servers.
  • Shlayer is a Trojan downloader and dropper for MacOS malware. It is primarily distributed through malicious websites, hijacked domains, and malicious advertising posing as a fake Adobe Flash updater. Note: Shlayer is the only malware of the top 10 that targets MacOS; the other 9 affect Microsoft Windows operating systems

Figure 1: Top 10 malware affecting SLTT educational institutions

Distributed Denial-of-Service Attacks

Cyber actors are causing disruptions to K-12 educational institutions—including third-party services supporting distance learning—with distributed denial-of-service (DDoS) attacks,  which temporarily limit or prevent users from conducting daily operations. The availability of DDoS-for-hire services provides opportunities for any motivated malicious cyber actor to conduct disruptive attacks regardless of experience level. Note: DDoS attacks overwhelm servers with a high level of internet traffic originating from many different sources, making it impossible to mitigate at a single source.

Video Conference Disruptions

Numerous reports received by the FBI, CISA, and MS-ISAC since March 2020 indicate uninvited users have disrupted live video-conferenced classroom sessions. These disruptions have included verbally harassing students and teachers, displaying pornography and/or violent images, and doxing meeting attendees (Note: doxing is the act of compiling or publishing personal information about an individual on the internet, typically with malicious intent). To enter classroom sessions, uninvited users have been observed:

  • Using student names to trick hosts into accepting them into class sessions, and
  • Accessing meetings from either publicly available links or links shared with outside users (e.g., students sharing links and/or passwords with friends).

Video conference sessions without proper control measures risk disruption or compromise of classroom conversations and exposure of sensitive information.

Additional Risks and Vulnerabilities

In addition to the recent reporting of distance learning disruptions received by the FBI, CISA, and MS-ISAC, malicious cyber actors are expected to continue seeking opportunities to exploit the evolving remote learning environment.

Social Engineering

Cyber actors could apply social engineering methods against students, parents, faculty, IT personnel, or other individuals involved in distance learning. Tactics, such as phishing, trick victims into revealing personal information (e.g., password or bank account information) or performing a task (e.g., clicking on a link). In such scenarios, a victim could receive what appears to be legitimate email that:

  • Requests personally identifiable information (PII) (e.g., full name, birthdate, student ID),
  • Directs the user to confirm a password or personal identification number (PIN),
  • Instructs the recipient to visit a website that is compromised by the cyber actor, or
  • Contains an attachment with malware.

Cyber actors also register web domains that are similar to legitimate websites in an attempt to capture individuals who mistype URLs or click on similar looking URLs. These types of attacks are referred to as domain spoofing or homograph attacks. For example, a user wanting to access could mistakenly click on (changed “o” to an “e”) or (changed letter “l” to a number “1”) (Note: this is a fictitious example to demonstrate how a user can mistakenly click and access a website without noticing subtle changes in website URLs). Victims believe they are on a legitimate website when, in reality, they are visiting a site controlled by a cyber actor.

Technology Vulnerabilities and Student Data

Whether as collateral for ransomware attacks or to sell on the dark web, cyber actors may seek to exploit the data-rich environment of student information in schools and education technology (edtech) services. The need for schools to rapidly transition to distance learning likely contributed to cybersecurity gaps, leaving schools vulnerable to attack. In addition, educational institutions that have outsourced their distance learning tools may have lost visibility into data security measures. Cyber actors could view the increased reliance on—and sharp usership growth in—these distance learning services and student data as lucrative targets.

Open/Exposed Ports

The FBI, CISA, and MS-ISAC frequently see malicious cyber actors exploiting exposed Remote Desktop Protocol (RDP) services to gain initial access to a network and, often, to manually deploy ransomware. For example, cyber actors will attack ports 445 (Server Message Block [SMB]) and 3389 (RDP) to gain network access. They are then positioned to move laterally throughout a network (often using SMB), escalate privileges, access and exfiltrate sensitive information, harvest credentials, or deploy a wide variety of malware. This popular attack vector allows cyber actors to maintain a low profile, as they are using a legitimate network service that provides them with the same functionality as any other remote user.

End-of-Life Software

End-of-Life (EOL) software is regularly exploited by cyber actors—often to gain initial access, deface websites, or further their reach in a network. Once a product reaches EOL, customers no longer receive security updates, technical support, or bug fixes. Unpatched and vulnerable servers are likely to be exploited by cyber actors, hindering an organization’s operational capacity.

Plans and Policies

The FBI and CISA encourage educational providers to maintain business continuity plans—the practice of executing essential functions through emergencies (e.g., cyberattacks)—to minimize service interruptions. Without planning, provision, and implementation of continuity principles, institutions may be unable to continue teaching and administrative operations. Evaluating continuity and capability will help identify potential operational gaps. Through identifying and addressing these gaps, institutions can establish a viable continuity program that will help keep them functioning during cyberattacks or other emergencies. The FBI and CISA suggest K-12 educational institutions review or establish patching plans, security policies, user agreements, and business continuity plans to ensure they address current threats posed by cyber actors.

Network Best Practices

  • Patch operating systems, software, and firmware as soon as manufacturers release updates.
  • Check configurations for every operating system version for educational institution-owned assets to prevent issues from arising that local users are unable to fix due to having local administration disabled.
  • Regularly change passwords to network systems and accounts and avoid reusing passwords for different accounts.
  • Use multi-factor authentication where possible.
  • Disable unused remote access/RDP ports and monitor remote access/RDP logs.
  • Implement application and remote access allow listing to only allow systems to execute programs known and permitted by the established security policy.
  • Audit user accounts with administrative privileges and configure access controls with least privilege in mind.
  • Audit logs to ensure new accounts are legitimate.
  • Scan for open or listening ports and mediate those that are not needed.
  • Identify critical assets such as student database servers and distance learning infrastructure; create backups of these systems and house the backups offline from the network.
  • Implement network segmentation. Sensitive data should not reside on the same server and network segment as the email environment.
  • Set antivirus and anti-malware solutions to automatically update; conduct regular scans.

User Awareness Best Practices

  • Focus on awareness and training. Because end users are targeted, make employees and students aware of the threats—such as ransomware and phishing scams—and how they are delivered. Additionally, provide users training on information security principles and techniques as well as overall emerging cybersecurity risks and vulnerabilities.
  • Ensure employees know who to contact when they see suspicious activity or when they believe they have been a victim of a cyberattack. This will ensure that the proper established mitigation strategy can be employed quickly and efficiently.
  • Monitor privacy settings and information available on social networking sites.

Ransomware Best Practices

The FBI and CISA do not recommend paying ransoms. Payment does not guarantee files will be recovered. It may also embolden adversaries to target additional organizations, encourage other criminal actors to engage in the distribution of ransomware, and/or fund illicit activities. However, regardless of whether your organization decided to pay the ransom, the FBI urges you to report ransomware incidents to your local FBI field office. Doing so provides the FBI with the critical information they need to prevent future attacks by identifying and tracking ransomware attackers and holding them accountable under U.S. law.

In addition to implementing the above network best practices, the FBI and CISA also recommend the following:

  • Regularly back up data, air gap, and password protect backup copies offline.
  • Implement a recovery plan to maintain and retain multiple copies of sensitive or proprietary data and servers in a physically separate, secure location.

Denial-of-Service Best Practices

  • Consider enrolling in a denial-of-service mitigation service that detects abnormal traffic flows and redirects traffic away from your network.
  • Create a partnership with your local internet service provider (ISP) prior to an event and work with your ISP to control network traffic attacking your network during an event.
  • Configure network firewalls to block unauthorized IP addresses and disable port forwarding.

Video-Conferencing Best Practices

  • Ensure participants use the most updated version of remote access/meeting applications.
  • Require passwords for session access.
  • Encourage students to avoid sharing passwords or meeting codes.
  • Establish a vetting process to identify participants as they arrive, such as a waiting room.
  • Establish policies to require participants to sign in using true names rather than aliases.
  • Ensure only the host controls screensharing privileges.
  • Implement a policy to prevent participants from entering rooms prior to host arrival and to prevent the host from exiting prior to the departure of all participants.

Edtech Implementation Considerations

  • When partnering with third-party and edtech services to support distance learning, educational institutions should consider the following:
  • The service provider’s cybersecurity policies and response plan in the event of a breach and their remediation practices:
    • How did the service provider resolve past cyber incidents? How did their cybersecurity practices change after these incidents?
  • The provider’s data security practices for their products and services (e.g., data encryption in transit and at rest, security audits, security training of staff, audit logs);
  • The provider’s data maintenance and storage practices (e.g., use of company servers, cloud storage, or third-party services);
  • Types of student data the provider collects and tracks (e.g., PII, academic, disciplinary, medical, biometric, IP addresses);
  • Entities to whom the provider will grant access to the student data (e.g., vendors);
  • How the provider will use student data (e.g., will they sell it to—or share it with—third parties for service enhancement, new product development, studies, marketing/advertising?);
  • The provider’s de-identification practices for student data; and
  • The provider’s policies on data retention and deletion.

Malware Defense

Table 1 identifies CISA-created Snort signatures, which have been successfully used to detect and defend against related attacks, for the malware variants listed below. Note: the listing is not fully comprehensive and should not be used at the exclusion of other detection methods.

Table 1: Malware signatures

Malware Signature
NanoCore alert tcp any any -> any $HTTP_PORTS (msg:"NANOCORE:HTTP GET URI contains 'FAD00979338'"; sid:00000000; rev:1; flow:established,to_server; content:"GET"; http_method; content:"getPluginName.php?PluginID=FAD00979338"; fast_pattern; http_uri; classtype:http-uri; metadata:service http;) 


alert tcp any any -> any $HTTP_PORTS (msg:"HTTP Client Header contains 'host|3a 20|'"; sid:00000000; rev:1; flow:established,to_server; flowbits:isnotset,<unique_ID>.tagged; content:"host|3a 20||0d 0a|"; http_header; fast_pattern:only; flowbits:set,<unique_ID>.tagged; tag:session,10,packets; classtype:http-header; metadata:service http;) 
Kovter alert tcp any any -> any $HTTP_PORTS (msg:"Kovter:HTTP URI POST to CnC Server"; sid:00000000; rev:1; flow:established,to_server; flowbits:isnotset,<unique_ID>.tagged; content:"POST / HTTP/1.1"; depth:15; content:"Content-Type|3a 20|application/x-www-form-urlencoded"; http_header; depth:47; fast_pattern; content:"User-Agent|3a 20|Mozilla/"; http_header; content:!"LOADCURRENCY"; nocase; content:!"Accept"; http_header; content:!"Referer|3a|"; http_header; content:!"Cookie|3a|"; nocase; http_header; pcre:"/^(?:[A-Za-z0-9+/]{4})*(?:[A-Za-z0-9+/]{2}==|[A-Za-z0-9+/]{3}=|[A-Za-z0-9+/]{4})$/P"; pcre:"/User-Agentx3a[^rn]+rnHostx3ax20(?:d{1,3}.){3}d{1,3}rnContent-Lengthx3ax20[1-5][0-9]{2,3}rn(?:Cache-Control|Pragma)x3a[^rn]+rn(?:rn)?$/H"; flowbits:set,<unique_ID>.tagged; tag:session,10,packets; classtype:nonstd-tcp; metadata:service http;)

alert tcp any any -> any $HTTP_PORTS (msg:"HTTP URI GET contains 'invoice_########.doc' (DRIDEX)"; sid:00000000; rev:1; flow:established,to_server; content:"invoice_"; http_uri; fast_pattern:only; content:".doc"; nocase; distance:8; within:4; content:"GET"; nocase; http_method; classtype:http-uri; metadata:service http;)
alert tcp any any -> any $HTTP_PORTS (msg:"HTTP Client Header contains 'Host|3a 20|tanevengledrep ru' (DRIDEX)"; sid:00000000; rev:1; flow:established,to_server; flowbits:isnotset,<unique_ID>.tagged; content:"Host|3a 20|tanevengledrep|2e|ru|0d 0a|"; http_header; fast_pattern:only; flowbits:set,<unique_ID>.tagged; tag:session,10,packets; classtype:http-header; metadata:service http;)


Advanced Persistent Threat Actors Targeting U.S. Think Tanks

This Advisory uses the MITRE Adversarial Tactics, Techniques, and Common Knowledge (ATT&CK®) framework. See the ATT&CK for Enterprise for all referenced threat actor tactics and techniques.

The Cybersecurity and Infrastructure Security Agency (CISA) and the Federal Bureau of Investigation (FBI) have observed persistent continued cyber intrusions by advanced persistent threat (APT) actors targeting U.S. think tanks. This malicious activity is often, but not exclusively, directed at individuals and organizations that focus on international affairs or national security policy.[1] The following guidance may assist U.S. think tanks in developing network defense procedures to prevent or rapidly detect these attacks.

APT actors have relied on multiple avenues for initial access. These have included low-effort capabilities such as spearphishing emails and third-party message services directed at both corporate and personal accounts, as well as exploiting vulnerable web-facing devices and remote connection capabilities. Increased telework during the COVID-19 pandemic has expanded workforce reliance on remote connectivity, affording malicious actors more opportunities to exploit those connections and to blend in with increased traffic. Attackers may leverage virtual private networks (VPNs) and other remote work tools to gain initial access or persistence on a victim’s network. When successful, these low-effort, high-reward approaches allow threat actors to steal sensitive information, acquire user credentials, and gain persistent access to victim networks.

Given the importance that think tanks can have in shaping U.S. policy, CISA and FBI urge individuals and organizations in the international affairs and national security sectors to immediately adopt a heightened state of awareness and implement the critical steps listed in the Mitigations section of this Advisory.

Click here for a PDF version of this report.

CISA and FBI recommend think tank organizations apply the following critical practices to strengthen their security posture.


  • Implement a training program to familiarize users with identifying social engineering techniques and phishing emails.


  • Log off remote connections when not in use.
  • Be vigilant against tailored spearphishing attacks targeting corporate and personal accounts (including both email and social media accounts).
  • Use different passwords for corporate and personal accounts.
  • Install antivirus software on personal devices to automatically scan and quarantine suspicious files.
  • Employ strong multi-factor authentication for personal accounts, if available.
  • Exercise caution when:
    • Opening email attachments, even if the attachment is expected and the sender appears to be known. See Using Caution with Email Attachments.
    • Using removable media (e.g., USB thumb drives, external drives, CDs).

IT Staff/Cybersecurity Personnel

  • Segment and segregate networks and functions.
  • Change the default username and password of applications and appliances.
  • Employ strong multi-factor authentication for corporate accounts.
  • Deploy antivirus software on organizational devices to automatically scan and quarantine suspicious files.
  • Apply encryption to data at rest and data in transit.
  • Use email security appliances to scan and remove malicious email attachments or links.
  • Monitor key internal security tools and identify anomalous behavior. Flag any known indicators of compromise or threat actor behaviors for immediate response.
  • Organizations can implement mitigations of varying complexity and restrictiveness to reduce the risk posed by threat actors who use Tor (The Onion Router) to carry out malicious activities. See the CISA-FBI Joint Cybersecurity Advisory on Defending Against Malicious Cyber Activity Originating from Tor for mitigation options and additional information.
  • Prevent exploitation of known software vulnerabilities by routinely applying software patches and upgrades. Foreign cyber threat actors continue to exploit publicly known—and often dated—software vulnerabilities against broad target sets, including public and private sector organizations. If these vulnerabilities are left unpatched, exploitation often requires few resources and provides threat actors with easy access to victim networks. Review CISA and FBI’s Top 10 Routinely Exploited Vulnerabilities and other CISA alerts that identify vulnerabilities exploited by foreign attackers.
  • Implement an antivirus program and a formalized patch management process.
  • Block certain websites and email attachments commonly associated with malware (e.g., .scr, .pif, .cpl, .dll, .exe).
  • Block email attachments that cannot be scanned by antivirus software (e.g., .zip files).
  • Implement Group Policy Object and firewall rules.
  • Implement filters at the email gateway and block suspicious IP addresses at the firewall.
  • Routinely audit domain and local accounts as well as their permission levels to look for situations that could allow an adversary to gain wide access by obtaining credentials of a privileged account.
  • Follow best practices for design and administration of the network to limit privileged account use across administrative tiers.
  • Implement a Domain-Based Message Authentication, Reporting & Conformance (DMARC) validation system.
  • Disable or block unnecessary remote services.
  • Limit access to remote services through centrally managed concentrators.
  • Deny direct remote access to internal systems or resources by using network proxies, gateways, and firewalls.
  • Limit unnecessary lateral communications.
  • Disable file and printer sharing services. If these services are required, use strong passwords or Active Directory authentication.
  • Ensure applications do not store sensitive data or credentials insecurely.
  • Enable a firewall on agency workstations, configured to deny unsolicited connection requests.
  • Disable unnecessary services on agency workstations and servers.
  • Scan for and remove suspicious email attachments; ensure any scanned attachment is its “true file type” (i.e., the extension matches the file header).
  • Monitor users’ web browsing habits; restrict access to suspicious or risky sites. Contact law enforcement or CISA immediately regarding any unauthorized network access identified.
  • Visit the MITRE ATT&CK techniques and tactics pages linked in the ATT&CK Profile section above for additional mitigation and detection strategies for this malicious activity targeting think tanks.