2022 · O - Public administration and defence · Cyber-espionage campaign via infected USB drives (removable-media malware)
Analysed 2026-11-06 by Hermes
Open this case in the tool →Cyber-espionage campaign via infected USB drives (removable-media malware)
Expertise to Execution
Evidence coverage: 2/6 stages, weight 15/60
Decisive stage: entry (clearest evidence) · Alternative: Strategy to Execution
(1) No source documents a victim-level policy at the 29 unnamed organisations, so the strategic expectation can only be reasoned: as multinational enterprises and government bodies operating IT in Africa, Europe, Asia and the US, the expectation was the known technical standard for removable media and endpoint hygiene - a standard the victim-side forensic source states explicitly, advising organisations to restrict access to external devices such as USB drives and scan devices before connecting them. (2) In practice, staff at at least 29 organisations plugged in infected USB drives - often picked up from shared public computers in internet cafes, print shops and a publicly accessible airport terminal in Harare - and clicked a disguised executable; the decade-old Sogu/Korplug/PlugX trojan then ran undetected, harvested and exfiltrated documents, copied itself onto every newly attached drive and spread machine to machine, in some cases crossing into air-gapped systems where it staged stolen data on the drive until that drive was plugged into an internet-connected machine; Mandiant, not the victims, surfaced the campaign. (3) The classification rests on the entry stage (the containment stage points to the same type): the knowledge sits with technical specialists - a known, published remedy against a more than decade-old malware family, stated by the firm investigating these very intrusions - and the chain broke in operations, because device restriction, media scanning and endpoint controls were absent in day-to-day practice at the affected organisations. Nothing in the sources shows leadership deciding, deferring or funding anything about removable media, which keeps the gap inside the material column (Expertise to Execution); because the victim organisations are unnamed, the origin/break reading rests on the operation-wide pattern documented by Mandiant's incident-response casework rather than an incident-specific victim confirmation, which caps confidence at 3. RM25 (Expertise to Execution lens): untrusted removable media flowed freely into the affected networks - drives infected on shared public computers were plugged into corporate and government systems and the malware even crossed into air-gapped machines - showing that the physical/media controls the security firm names as the remedy (restricting external devices, scanning before connection) were not implemented in operations, which is what enabled the intrusion (causal role: Enabling). RM28 (Expertise to Execution lens): a more than decade-old, extensively documented trojan family executed on victim endpoints, persisted under names masquerading as security products and spread drive to drive without being blocked, so endpoint malware protection and application control were absent or ineffective against a family that was anything but a novel evasion (causal role: Enabling).
RM25 — Physical / media security gapRM28 — Endpoint / malware-protection gapVictim organisations are unnamed by design; the classification covers the described population of 29 organisations and the mechanism is operation-wide (Mandiant incident-response casework), not confirmed for an individual victim - confidence capped at 3 accordingly. Actor naming: the case file and Wired use UNC53 (mWise disclosure, 19-09-23); Mandiant's July 2023 Managed Defense report on the same SOGU USB campaign attributes it to TEMP.Hex (Campaign 22-054). Data quality: the case file's second source_url (Wired Kia/Hyundai weekly roundup) only mentions the main article; the analysis rests on the main article itself. Country left null: multi-country campaign (receiver_country lists Kenya, Egypt, Zimbabwe, Madagascar, Tanzania, Ghana, the United States plus Asia and Europe regions) with no single victim country. Sector: diffuse multi-sector population (receiver_subcategory: government ministries, health, finance, plus unnamed corporate targets; Wired adds consulting, marketing, engineering, construction, mining, education, pharmaceuticals); coded from the row's leading receiver category (State institutions / political system), matching the convention used for diffuse campaigns in this run. Detection, egress, recovery and response stages are recorded as gaps not documented: no per-victim dwell time, no victim statements, disclosure was by the security company. EuRepoC impact coding: minor data breach/exfiltration of non-sensitive commercial information, no disruption.