2020 · O - Public administration and defence · Credential-harvesting phishing campaign against government e-mail portals
Analysed 2026-11-06 by Hermes
Open this case in the tool →Credential-harvesting phishing campaign against government e-mail portals
Expertise to Execution
Evidence coverage: 2/6 stages, weight 12/60
Decisive stage: detection (clearest evidence) · Alternative: Strategy to Execution
(1) The expectation at the victims - government ministries running webmail portals for foreign-affairs, finance, defence and intelligence staff - was that portal authentication is protected and that infrastructure impersonating the organisation's own mail login pages is detected and removed; no victim policy artefact is documented, so the expectation is derived from the targets' criticality. (2) In practice, an actor maintained 50+ look-alike 'mail.<real-domain>' harvesting pages against ministries in at least eleven countries from spring 2020 until Cyjax's analysts disclosed the campaign in September 2021, with 15 pages still live at that date; the aimed-at effect is government e-mail account access, and EuRepoC scores the incident as data theft plus hijacking/system misuse. (3) This is a knowing-doing gap inside the material column: the remedies that defeat portal credential harvesting (MFA and credential safeguards on the portals, look-alike-domain and credential monitoring, takedown) are standard technical practice needing no strategic decision, yet the infrastructure ran unnoticed by every targeted government for about seventeen months and was found only by an outside firm; the classification rests on the detection stage, with the entry stage (successful credential capture at cloned portals) pointing to the same type. The distribution vector is not documented (Cyjax found no phishing e-mails) and is not load-bearing. Attribution is tentative and campaign-level, and no individual victim confirmed compromise, so the origin/break reading is coded as inference. (4) RM34 (Expertise to Execution lens): the targeted governments' detection tooling and processes were insufficiently configured, staffed or tuned to catch a large look-alike harvesting campaign against their own mail portals - a ~17-month dwell with 15 pages still live at external discovery (causal role: Obscuring). RM22 (Expertise to Execution lens): identity controls that would neutralise harvested portal passwords, above all MFA and leaked-credential monitoring, were evidently not realised on the targeted government webmail portals, letting the attacker collect credentials usable to log in as legitimate users (causal role: Enabling).
RM22 — Authentication / credential exposureRM34 — Monitoring / detection / alert-handling gapCampaign-level EuRepoC row: the victim is a diffuse set of government bodies in ~10-12 countries (receiver_country lists Uzbekistan, Azerbaijan, Belarus, China, Georgia, Kyrgyzstan, Pakistan, Russia, Turkey, Ukraine; Cyjax's IOC list adds Armenia and Turkmenistan), so Country is left blank per the multi-organisation-campaign convention. Year 2020 = eurepoc_extra.start_date and the campaign start Cyjax dates to spring 2020; disclosure was September 2021. Attribution is tentative: initiator Unknown, state-affiliated per the dataset; the UNC1151 / Hades (Sandworm) link is Cyjax's own hypothesis via one OVH IP shared with Operation TrickyMouse - unverified. Data quality: the receiver_category and receiver_subcategory fields repeat an identical per-country block about ten times (an evident data-quality artefact). RM24 (human-factor) was considered but not coded: no source documents a victim-side training or verification failure, and Cyjax did not even uncover the phishing e-mails, so its exclusion condition is not cleared. Unattended run: nothing required consent beyond permitted web research.