Join our Newsletter — 33% off our NHI Course
Home› Guides› Winter Vivern Threat Actor Profile Guide: webmail XSS…
Guide Threat Intelligence & Threat Actors

Winter Vivern Threat Actor Profile Guide: webmail XSS espionage and how to defend

PDF for newsletter subscribers · free to read online

← All guides
By NHI Mgmt Group, edited by Lalit Choda Updated 27 October 2026 10 min read
On this page

Winter Vivern reads government email by turning the webmail portal against its own users: a single crafted message, opened in a browser, is enough to run its JavaScript inside a Roundcube or Zimbra session. MITRE ATT&CK tracks the group as G1035 and describes it as linked to Russian and Belarusian interests, active since at least 2020 against European government and NGO entities with sporadic targeting in India and the United States. ESET documented its use of a Roundcube zero-day, CVE-2023-5631, in October 2023, and Proofpoint, which tracks the group as TA473, documented exploitation of a Zimbra cross-site scripting flaw against NATO-aligned governments from February 2023. This profile is for email and web platform owners, SOC analysts and threat intelligence teams in government, diplomatic and policy organisations. It covers names, targeting, ATT&CK techniques, how the group steals credentials and sessions, a dated incident list and a detection and hardening plan. The wider Russian state picture is in the Russia State Cyber Operations Guide.

Key takeaways

  • Webmail is the primary target. The group has exploited Roundcube (CVE-2020-35730 and the zero-day CVE-2023-5631) and Zimbra (CVE-2022-27926) to run script in victims' mail sessions.
  • Opening the message is enough. ESET found the CVE-2023-5631 exploit needed no interaction beyond viewing the email in a browser; there was no attachment to block.
  • The goal is mailbox access and credentials. Payloads listed folders and exfiltrated messages, and Proofpoint saw portal-specific JavaScript capture usernames, passwords, CSRF tokens and session tokens.
  • Old-fashioned lures still run alongside. Fake antivirus scanners, cloned ministry websites and macro documents lead to PowerShell scripts, scheduled tasks and file theft over HTTP.
  • Patch speed and portal exposure decide the outcome. Proofpoint's advice was to patch Zimbra on public webmail portals and restrict their exposure to the open internet; ESET's was to upgrade Roundcube immediately.

Who they are (names and aliases)

NameUsed byNotes
Winter VivernDomainTools (2021), MITRE ATT&CK (G1035), ESET, SentinelOneThe most widely used name.
TA473ProofpointTracked by Proofpoint since at least 2021.
UAC-0114CERT-UAUkraine's CERT designation, used in early 2023 reporting.

Attribution as stated: ATT&CK says the group is linked to Russian and Belarusian interests. SentinelOne describes it as operating with pro-Russian objectives closely aligned with the interests of the Belarusian and Russian governments, and as resource-limited but creative. Proofpoint agrees the targeting broadly supports Russian and Belarusian goals regarding the war in Ukraine. None of these sources names a specific government agency as the operator, so neither should your reporting.

Targeting and motivation

  • Government and diplomacy: European government, military and diplomatic organisations, with a particular focus on the public webmail portals of NATO-aligned governments (Proofpoint). SentinelOne names the foreign ministries of Ukraine and Italy and government agencies in Poland.
  • Think tanks and NGOs: ESET's 2023 Roundcube victims were governmental entities and a think tank in Europe; ATT&CK lists NGOs.
  • People who matter to policy: Proofpoint observed phishing against US elected officials and staff in late 2022, and targets who are experts in European politics or the economics of the Russia and Ukraine conflict.
  • Telecoms and India: SentinelOne highlights private telecommunications organisations supporting Ukraine and users of the Indian government email service email.gov.in, phished in mid-2022.
  • Motivation: espionage, specifically the content of official mailboxes and the credentials that unlock them.

Tactics and techniques (MITRE ATT&CK)

Mapping from ATT&CK v20 group G1035 procedure descriptions. IDs checked against the v20 dataset.

TacticTechniqueObserved behaviour
ReconnaissanceT1595.002 Vulnerability ScanningRemotely hosted Acunetix instances used to find unpatched webmail portals.
Resource developmentT1583.001 Domains; T1584.006 Web ServicesLook-alike domains for government entities; compromised WordPress sites hosting payloads.
Initial accessT1190 Exploit Public-Facing ApplicationRoundcube and Zimbra XSS flaws; ATT&CK also records use of the Follina vulnerability (CVE-2022-30190).
Initial accessT1566.001 Spearphishing Attachment; T1189 Drive-by CompromiseMalicious documents; cloned government websites delivering fake antivirus software.
ExecutionT1059.007 JavaScript; T1059.001 PowerShell; T1059.003 Windows Command Shell; T1204.001 Malicious LinkJavaScript delivered through webmail XSS; macros passing execution to PowerShell; batch scripts disguised as virus scanners.
PersistenceT1053.005 Scheduled TaskPowerShell creates scheduled tasks that periodically fetch remote payloads.
StealthT1036.004 Masquerade Task or Service; T1140 Deobfuscate/Decode Files or InformationTools posing as virus scanners; base64-encoded exploit payloads in emails.
Credential accessT1056.003 Web Portal CapturePages mimicking government email logon sites to collect credentials.
CollectionT1114.001 Local Email Collection; T1083 File and Directory Discovery; T1113 Screen Capture; T1119 Automated CollectionJavaScript lists folders and emails on the mail server and exfiltrates messages; scripts take screenshots and recursively collect files.
Exfiltration and C2T1041 Exfiltration Over C2 Channel; T1020 Automated Exfiltration; T1071.001 Web ProtocolsHTTP and HTTPS for command and control and exfiltration.

How they abuse identities and credentials

For Winter Vivern the mailbox session is the identity. A cross-site scripting payload runs with the same authority as the logged-in user, so MFA at sign-in does not stop it: the session is already authenticated.

  • Session and token theft: Proofpoint's analysis of the Zimbra payloads found JavaScript written for each target portal that stole CSRF tokens and active session tokens from cookies or responses, and in some cases Roundcube request tokens. In defender terms this is the behaviour described by ATT&CK T1539 Steal Web Session Cookie. See the Session Hijacking and Cookie Theft Guide.
  • Credential capture in the real portal: the same payloads captured usernames and passwords entered on the legitimate login page, then tried to log in with them, and showed fake POP3 or IMAP setup pages. Because the script runs on the genuine domain, URL checks by the user do not help.
  • Look-alike logon pages: separately, the group built phishing pages imitating government email sign-in (T1056.003), including against email.gov.in users (SentinelOne).
  • Trusted sender abuse: Proofpoint saw emails sent from compromised accounts and spoofed to look like a colleague or a peer government, carrying a legitimate-looking URL hyperlinked to actor infrastructure.
  • Mailbox content as the prize: the CVE-2023-5631 payload could list folders and emails and send messages to the attacker's server (ESET), which is collection without ever needing the password.

Defensive consequence: password resets alone are not enough after a webmail XSS compromise. Invalidate all sessions, rotate application-specific passwords and IMAP or POP credentials, and review forwarding rules and connected clients. Phishing-resistant MFA still matters because it defeats the look-alike page route; see the Phishing-Resistant MFA Guide.

Notable incidents

WhenEventSource
2021DomainTools reports re-crafted government macro documents and names the group.DomainTools, cited by ATT&CK
Mid-2022Credential phishing pages target users of India's email.gov.in.SentinelOne
Late 2022Phishing against US elected officials and staff.Proofpoint
Early 2023Campaigns against Ukrainian and Italian foreign ministries and Polish government agencies; fake scanner lures and the APERETIF trojan reported by CERT-UA (UAC-0114) and Poland's CBZC.SentinelOne, March 2023
From February 2023Exploitation of Zimbra CVE-2022-27926 against NATO-aligned government webmail. CISA added the CVE to KEV on 3 April 2023.Proofpoint; CISA KEV
August to September 2023Exploitation of Roundcube CVE-2020-35730, a flaw also used by Sednit (APT28). CISA added it to KEV on 22 June 2023.ESET; CISA KEV
11 October 2023Zero-day exploitation of Roundcube CVE-2023-5631 against European government entities and a think tank; fixed in Roundcube 1.6.4, 1.5.5 and 1.4.15 on 16 October 2023; KEV listing 26 October 2023.ESET; CISA KEV

Detection and defence

Webmail platform

  1. Patch inside days, not cycles. Subscribe to Roundcube and Zimbra security announcements and track the KEV entries above. ESET notes CVE-2023-5631 affected fully patched servers at discovery, so patching must be paired with exposure reduction. See the Roundcube Webmail Hardening Guide and the Zimbra Collaboration Hardening Guide.
  2. Reduce exposure. Proofpoint recommends restricting public webmail portals from the open internet: put them behind a VPN, zero trust access proxy or at minimum geographic and IP restrictions for sensitive mailboxes.
  3. Contain script. A strict Content Security Policy that blocks inline script and limits script sources to the webmail origin makes XSS payloads that fetch second stages from attacker domains much harder to run. Test in report-only mode first. See the Content Security Policy Guide.
  4. Harden cookies. HttpOnly, Secure and SameSite on session cookies limit what script can read and send. See the Cookie Security Guide.

What to log and hunt

SignalSourceLinked behaviour
Inbound HTML mail with SVG elements, onerror handlers or nested base64 contentMail gateway, server message storeCVE-2023-5631 trigger (ESET); nested base64 (Proofpoint)
Browser requests from the webmail origin to newly registered or unknown domainsProxy, DNS logs, CSP violation reportsSecond-stage loader fetch
Bulk folder listing or message retrieval in a short window by one sessionWebmail application and IMAP logsT1114.001 email collection
The same session token used from two IP addresses or user agentsWebmail access logsSession token theft
Login with correct credentials from new infrastructure shortly after a user opened a suspicious mailAuthentication logsCaptured credential replay
Scheduled tasks created by PowerShell that download from remote URLsWindows Event ID 4698, EDRT1053.005, T1105
Vulnerability scanner fingerprints (for example Acunetix) against webmail hostsWAF and web server logsT1595.002 reconnaissance

Response

  • Scope the mail store: search all mailboxes for the malicious message, because every recipient who opened it in a browser may be affected.
  • Kill sessions and rotate: invalidate all webmail sessions, reset passwords and application passwords, and review forwarding rules, filters and identities added to the account.
  • Treat it as a data loss event: assume opened mailboxes were read and exfiltrated; follow the Phishing Incident Response Playbook and preserve server logs for the Email Forensics and Header Analysis Guide workflow.

Framework mapping: patching maps to NIST SP 800-53 SI-2 (Flaw Remediation); exposure reduction to SC-7 (Boundary Protection); script controls to SC-18 (Mobile Code); monitoring to SI-4 and CSF 2.0 DE.CM-09.

Metrics: days from a Roundcube or Zimbra security release to production patch (target under 7 for internet-facing portals); percentage of webmail portals reachable only through an access gateway; CSP violation reports from the webmail origin investigated within 24 hours; time from detection to global session revocation for affected users.

Practitioner checklist

  • List every Roundcube and Zimbra instance, with version, plugins and whether it is reachable from the internet.
  • Patch CVE-2023-5631, CVE-2020-35730 and CVE-2022-27926 and set a 7-day target for future webmail security releases.
  • Move sensitive webmail behind VPN or an identity-aware access proxy.
  • Deploy a strict Content Security Policy on webmail, starting in report-only mode, and collect violation reports.
  • Set HttpOnly, Secure and SameSite on session cookies and bind sessions to client context where the platform allows.
  • Alert on bulk folder and message access per session and on session reuse from new IPs.
  • Enrol diplomatic and policy staff in phishing-resistant MFA and brief them on look-alike ministry pages and fake antivirus lures.
  • Pre-write the response: mailbox-wide search, global session revocation, credential and application password rotation.

Standards and references

Related NHI Mgmt Group resources: Threat Actor Groups Guide · Roundcube Webmail Hardening Guide · Cross-Site Scripting (XSS) Prevention Guide · Email Security Guide

PDF for subscribers

Take this guide with you

Every section, table and checklist in one PDF, with links to the sources and related guides. Subscribe to our weekly NHI & AI Identity Journal and we’ll email you a download link.

Free weekly newsletter

Subscribe to the NHI & AI Identity Journal

The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.

Bonus 33% off our NHI Course when you subscribe.

NHIMG Editorial Note
Researched and written by the NHI Mgmt Group editorial team using AI tools, under the editorial direction of Lalit Choda. Facts are drawn from the sources listed on this page and checked by our automated quality process. Last updated 27 October 2026. Spotted an error? Use our contact page.
NHI Mgmt Group, the independent authority on Non-Human Identity, IAM and Agentic AI security. nhimg.org

PDF Downloads for Newsletter Subscribers

Get the guide

Enter your details and we’ll email you a download link.

The full guide is free to read online. The PDF is a thank you for subscribing.

Fill in all fields marked * and tick the PDF box to continue.