Sandworm APT's initial access subgroup hits organizations accross the globe
Vulnerabilities mentionedAll →
| CVE | Vulnerability | CVSS | EPSS | Flags | Affected | Exposure | Published |
|---|---|---|---|---|---|---|---|
| CVE-2021-34473 | SSRF (CWE-918) in Microsoft Exchange Server Enabling RCE (ProxyShell) Microsoft Exchange Server contains a server-side request forgery (SSRF, CWE-918) in the Autodiscover service that unauthenticated remote attackers can reach over HTTP. This flaw is the first stage of the widely documented 'ProxyShell' chain, in which the SSRF is combined with privilege escalation and an arbitrary file write in the Exchange PowerShell backend to achieve unauthenticated remote code execution on the server. An attacker who successfully exploits it gains the ability to run code on the Exchange server, and ransomware operators (e.g., Cuba ransomware) leveraged this chain to deploy payloads. Organizations running on-premises Microsoft Exchange Server (2013, 2016, and 2019 per the vendor's advisories) are affected; hosted Exchange Online is a separately managed cloud service. Exploitation is confirmed in the wild: CISA added the flaw to the KEV catalog on 2021-11-03 with known ransomware use, and EPSS assigns it a 100% probability of exploitation within 30 days (100th percentile). Do: Apply Microsoft's July 2021 (or later) Exchange Server security updates, which include the ProxyShell fixes, to all on-premises servers. If patching is delayed, restrict internet access to the Autodiscover/EWS frontend endpoints and consider the URL Rewrite mitigation Microsoft published for ProxyShell. Given the KEV listing and known ransomware use, treat internet-facing, unpatched Exchange servers as potentially compromised and hunt for webshells, unusual Exchange processes, and suspicious account or mailbox activity. | 9.1 | 100% | KEV ransomware PoC |
| masshundreds of thousands of on-prem Exchange servers; public internet scans at the time of disclosure showed roughly 400,000+ internet-exposed Exchange instances,… | |
| CVE-2022-41352 | Path Traversal Arbitrary File Upload in Synacor Zimbra Collaboration Suite CVE-2022-41352 is an unauthenticated arbitrary file upload flaw in Zimbra Collaboration Suite (ZCS) 8.8.15 and 9.0, rooted in a path traversal weakness (CWE-22) in how the amavis mail scanner hands archives to the cpio utility. An attacker sends a specially crafted archive via email; cpio follows traversal entries and extracts attacker-controlled files into the web root at /opt/zimbra/jetty/webapps/zimbra/public. Uploaded files in that web-accessible directory can lead to remote code execution and incorrect access to other users' accounts, giving network-level attackers high impact on confidentiality, integrity, and availability (CVSS 9.8). Any ZCS 8.8.15 or 9.0 deployment that processes mail with cpio is affected, particularly on Red Hat/CentOS 7 and later where the pax utility (which Zimbra recommends as the safe alternative) is not installed by default. The flaw is actively exploited in the wild: it was added to CISA's Known Exploited Vulnerabilities Catalog on 2022-10-20 with known ransomware use, carries a 95.5% EPSS score, and public reporting ties Zimbra exploitation to campaigns including Russia-linked Sandworm/Seashell Blizzard access operations. Do: Apply updates per vendor instructions, and as an immediate mitigation install the pax package on the Zimbra host (amavis automatically prefers pax over cpio once it is present; note pax is not in default Red Hat installations after RHEL/CentOS 6, while it is a prerequisite on Ubuntu). Also inspect /opt/zimbra/jetty/webapps/zimbra/public for unexpected uploaded files and review mail flow logs for suspicious archive attachments, given confirmed in-the-wild and ransomware-related exploitation. | 9.8 | 95% | KEV ransomware PoC |
| largetens of thousands of internet-exposed Zimbra servers (public scan counts have shown roughly 40,000-60,000 exposed instances) | |
| CVE-2023-23397 | Zero-Click Elevation of Privilege in Microsoft Outlook (Forced NTLM Credential Leak) CVE-2023-23397 is an elevation of privilege vulnerability in Microsoft Outlook caused by improper input validation (CWE-20) combined with authentication bypass via spoofed authentication data on the channel (CWE-294), allowing an attacker to force Outlook to authenticate to an attacker-controlled SMB/WebDAV server. It is triggered when Outlook processes a crafted email or calendar object — for example a meeting or task reminder whose sound property points to an attacker-supplied UNC path — and requires no user interaction. That authentication exchange leaks the victim's NTLM credential hash, which the attacker can crack offline or relay to authenticate as the victim and access resources such as Exchange mailboxes, effectively escalating privileges. Affected software spans Microsoft 365 Apps, Microsoft Office (including the Long Term Servicing Channel), and Microsoft Outlook, which are deployed across enterprises, governments, and militaries worldwide. It is actively exploited in the wild — added to CISA's Known Exploited Vulnerabilities catalog on 2023-03-14 with a 97.4% EPSS — and Microsoft has warned of exploitation by Russia-aligned threat actors in campaigns against government and military mail servers, with patches shipped in Microsoft's March 2023 security updates. Do: Apply Microsoft's March 2023 security updates to Microsoft 365 Apps, Office/LTSC, and Outlook immediately, per CISA's required action. As interim mitigation, enable Extended Protection for Authentication or add accounts to the Protected Users group to block the NTLM credential leak, and audit calendar and task reminder sound properties for UNC paths (Microsoft published an audit/cleanup script for this) while watching for unexpected outbound SMB/WebDAV connections from hosts running Outlook. | 9.8 | 97% | KEV |
| masson the order of hundreds of millions of users (Outlook ships with Microsoft Office/Microsoft 365, the dominant enterprise and government email suite) | |
| CVE-2023-32315 | Unauthenticated Path Traversal in Ignite Realtime Openfire Admin Console CVE-2023-32315 is an unauthenticated path traversal flaw (CWE-22) in the Ignite Realtime Openfire XMPP server that allows a remote attacker to reach pages of the Openfire Admin Console that are reserved for administrative users. It is triggered by sending a crafted HTTP request to the admin console web interface containing path traversal sequences, which bypasses the authentication check protecting those restricted pages. Successful exploitation exposes administrative console functionality to an unauthenticated attacker, potentially enabling further compromise of the server and the chat environment it hosts. Any organization running Openfire is affected, particularly instances whose admin console interface is reachable from the internet. The flaw was added to CISA's Known Exploited Vulnerabilities catalog on 2023-08-24 and carries an EPSS probability of 100% (top percentile), indicating active exploitation in the wild; ransomware use is not yet documented. Do: Upgrade Openfire to the patched release recommended in the Ignite Realtime advisory for CVE-2023-32315, per the CISA KEV required action; if patching must be delayed, restrict the admin console (default ports 9090/9091) to trusted networks or place it behind a VPN. Review access logs for requests to admin console pages containing traversal sequences, and check for unauthorized admin accounts or configuration changes, since KEV listing confirms exploitation in the wild. | 7.5 | 100% | KEV PoC ×2 |
| large≈ tens of thousands of internet-exposed Openfire servers (~30,000–40,000 per public scans) | |
| CVE-2023-42793 | Authentication bypass in JetBrains TeamCity enables unauthenticated RCE JetBrains TeamCity Server, a widely used CI/CD build server, contains an authentication bypass vulnerability (CWE-288) that lets a remote, unauthenticated attacker gain administrative access without valid credentials. By sending crafted requests to the TeamCity server over the network, the attacker bypasses authentication and can then execute arbitrary code on the server via administrative and build features, achieving full remote code execution. An attacker gains control of the build server and, with it, access to source code, build artifacts, stored secrets and credentials, and a foothold for lateral movement or ransomware deployment. Any organization running an affected TeamCity Server is affected, especially instances reachable from the internet. Exploitation is confirmed in the wild: CISA added the flaw to its KEV catalog on 2023-10-04 with known ransomware use, and EPSS assigns a 100% probability of exploitation within 30 days. Do: Upgrade TeamCity Server to 2023.05.4 or later per JetBrains' instructions, or apply vendor mitigations or discontinue use if patching is not possible (per the CISA KEV required action). Also take unpatched instances off the public internet, and hunt for signs of compromise such as unauthorized administrator accounts, unexpected changes in audit logs and build configurations, and stored secrets or tokens that may have been stolen, given known ransomware exploitation. | 9.8 | 100% | KEV ransomware PoC ×2 |
| largeTens of thousands of TeamCity Server deployments, of which several thousand are internet-exposed | |
| CVE-2023-48788 | Unauthenticated SQL Injection in Fortinet FortiClient EMS Fortinet FortiClient EMS — the central management server for FortiClient endpoint deployments — contains a SQL injection flaw (CWE-89) in versions 7.0.1 through 7.0.10 and 7.2.0 through 7.2.2. The CVSS 3.1 vector (AV:N/AC:L/PR:N/UI:N) shows it can be triggered remotely with no credentials and no user interaction: an unauthenticated attacker sends specially crafted packets to the vulnerable management server and can execute unauthorized code or commands. Successful exploitation effectively yields remote code execution on the EMS server and access to its database, enabling follow-on actions such as credential theft, abuse of endpoint management functions, and ransomware deployment. Any organization running the affected EMS versions is exposed, especially where the management server is reachable from the internet. Exploitation is confirmed in the wild: CISA added the bug to the KEV catalog on 2024-03-25 with known ransomware use, and EPSS assigns a ~98.4% probability of exploitation within 30 days (100th percentile). Do: Upgrade FortiClient EMS to the fixed releases per Fortinet's advisory for this CVE (7.2.3 and 7.0.11 or later, i.e., beyond the 7.2.2 and 7.0.10 affected ranges); the CISA KEV required action is to apply vendor mitigations or discontinue use if mitigations are unavailable. Until patched, limit exposure of the EMS web interface to untrusted networks and hunt for signs of compromise — anomalous requests to the management console, unexpected database or admin activity, and follow-on ransomware behavior — since exploitation with known ransomware use is confirmed. | 9.8 | 98% | KEV ransomware |
| largetens of thousands of EMS deployments worldwide, with a smaller subset (likely thousands) internet-exposed | |
| CVE-2024-1709 | Authentication Bypass in ConnectWise ScreenConnect Creates Rogue Admin Accounts ConnectWise ScreenConnect (ConnectWise Control), a widely used remote-access and remote-monitoring tool, contains an authentication bypass (CWE-288) in its management interface. An attacker needs only network access to the management interface to trigger the flaw, with no valid credentials or user interaction required. A successful attacker gains administrative control of the ScreenConnect server by creating a new administrator-level account, providing a foothold that has already been used in ransomware campaigns against downstream managed environments. Any organization running ConnectWise ScreenConnect is affected, especially managed service providers and IT teams whose management interface is reachable from the internet; the source data specifies affected products but no version ranges. Exploitation is confirmed and urgent: CISA added the flaw to the KEV on 2024-02-22 with known ransomware use, EPSS assigns a 100% probability of exploitation within 30 days, and ConnectWise warned that no patch was available at the time of disclosure. Do: Follow ConnectWise's instructions immediately: no patch existed at disclosure, so apply the vendor's mitigations or, per the CISA KEV required action, restrict internet exposure of the management interface or discontinue use until mitigations are available, then upgrade to the vendor's patched release as soon as it ships. Audit ScreenConnect servers for unexpectedly created administrator-level accounts and unusual remote sessions, which are the attack's artifacts. Prioritize any instance whose management interface is reachable from the internet, given confirmed in-the-wild exploitation and known ransomware use. | 10.0 | 100% | KEV ransomware PoC ×3 |
| masstens of thousands of internet-exposed ScreenConnect servers (on the order of 10,000-30,000 instances in public internet scans at disclosure), managing millions… |
Full article683 words · extracted from helpnetsecurity.com · click to collapse
A subgroup of Russia’s Sandworm APT has been working to achieve initial and persistent access to the IT networks of organizations working in economic sectors Russia is interested in.
“In 2022, its primary focus was Ukraine, specifically targeting the energy, retail, education, consulting, and agriculture sectors. In 2023, it globalized the scope of its compromises, leading to persistent access within numerous sectors in the United States, Europe, Central Asia, and the Middle East,” Microsoft’s researchers have shared on Wednesday.
“In 2024, while the exposure of multiple vulnerabilities likely offered the subgroup more access than ever, it appeared to have honed its focus to the United States, Canada, Australia, and the United Kingdom.”

The subgroup’s targets span the world (Source: Microsoft)
About Sandworm
Sandworm (per Microsoft: “Seashell Blizzard”) is a threat group that has been associated by researchers to the Russian Military Intelligence Unit 74455 (GRU).
The group has been involved in destructive attacks such as KillDisk and NotPetya.
In 2020, the US unsealed an indictment against six GRU officers believed to be Sandworm members.
The tactics, techniques, and procedures (TTPs) of the Sandworm initial access subgroup
The subgroup uses public scan databases and focuses on exploiting targets’ vulnerable Internet-facing infrastructure. They have been known to leverage the following vulnerabilities:
- CVE-2021-34473 (in Microsoft Exchange)
- CVE-2022-41352 (in Zimbra Collaboration)
- CVE-2023-32315 (in OpenFire)
- CVE-2023-42793 (in JetBrains TeamCity)
- CVE-2023-23397 (in Microsoft Outlook)
- CVE-2024-1709 (in Connectwise ScreenConnect)
- CVE-2023-48788 (in Fortinet FortiClient EMS), and
- An unknown vulnerability in the JBoss Enterprise Application Platform
For persistence, they initially deployed web shells. In early 2024, they also began installing and leveraging legitimate remote monitoring and management (RMM) tools like Atera Agent and Splashtop Remote Services.
The RMM tools allowed them to deploy secondary tools for stealing and exfiltrating credentials.
The subgroup’s operational lifecycle (Source: Microsoft)
“Among a subgroup of victims, Seashell Blizzard carried out unique post-compromise activity, indicating that the threat actor sought more durable persistence and direct access. In these cases, Seashell Blizzard deployed OpenSSH with a unique public key, allowing them to access compromised systems using an actor-controlled account and credential,” Microsoft says.
For added persistence, they also registered compromised systems as a Tor hidden service (a technique Microsoft dubbed ShadowLink).
“Systems compromised with ShadowLink receive a unique .onion address, making them remotely accessible via the Tor network. This capability allows Seashell Blizzard to bypass common exploit patterns of deploying a RAT, which commonly leverages some form of C2 to actor-controlled infrastructure that are often easily audited and identified by network administrators,” the researchers explained.
“Instead, by relying on Tor hidden services, the compromised system creates a persistent circuit to the Tor network, acting as a covert tunnel, effectively cloaking all inbound connections to the affected asset and limiting exposures from both the actor and victim environment.”
The group also modified network resources such as Outlook Web Access (OWA) sign-in pages (to collect credentials for lateral movement) and DNS configurations (possibly to intercept credentials from critical authentication services).
Setting the stage for future attacks
The subgroup has likely used a “spray and pray” approach to achieving compromises and has also compromized organizations that have limited or no utility to Russia’s strategic interests, according to Microsoft researchers. “In cases where a strategically significant target is compromised, we have observed significant later post-compromise activity.”
Among the actual targets were organizations in the energy, oil and gas, telecommunications, shipping, and arms manufacturing sectors, as well as international governments.
“This subgroup, which is characterized within the broader Seashell Blizzard organization by its near-global reach, represents an expansion in both the geographical targeting conducted by Seashell Blizzard and the scope of its operations. At the same time, Seashell Blizzard’s far-reaching, opportunistic access methods likely offer Russia expansive opportunities for niche operations and activities that will continue to be valuable over the medium term,” the researchers pointed out.
Microsoft has released indicators of compromise, mitigation and protection guidance, detections and alerts pointing to possible Seashell Blizzard activity, as well as threat hunting queries.

Subscribe to our breaking news e-mail alert to never miss out on the latest breaches, vulnerabilities and cybersecurity threats. Subscribe here!

Text extracted automatically; images, tables and formatting may be missing. Original: https://www.helpnetsecurity.com/2025/02/13/sandworm-apts-initial-access-subgroup-hits-organizations-accross-the-globe/