2017 · O - Public administration and defence · State-sponsored cyber-espionage campaign (data theft and system hijacking)
Analysed 2026-11-06 by Hermes
Open this case in the tool →State-sponsored cyber-espionage campaign (data theft and system hijacking)
No DV evidenced
Alternative: none defensible
(1) The row aggregates APT37's 2017 expansion of targeting beyond the Korean peninsula to Japan, Vietnam and the Middle East across government, transport, health, chemicals, telecommunications, finance, defence-industry and science targets; the strategic expectation the DV framework would need - a named victim organisation's own policy, standard or strategy - is nowhere stated, because the case is a campaign description, not an incident at one organisation. (2) What the record documents is the actor: FireEye's February 2018 special report (the dataset's own source) sets out APT37's targeting profile and mission, its infection vectors (targeted spear phishing with exploit documents such as CVE-2017-0199 and the CVE-2018-4878 Flash zero-day, strategic web compromises of South Korean sites, torrent distribution), its malware suite and exfiltration tools, and its attribution to North Korea with high confidence; secondary reporting (Bleeping Computer, 21 February 2018) confirms only this actor-side picture, including attacks on a Vietnamese trading/transport executive and a Middle Eastern financial company after business deals with North Korea went bad. (3) The DV is a property of a victim organisation's situation - where its doctrinal expectations or its technical expertise failed to reach execution or leadership; here no victim organisation is named and no source makes any victim's strategy, expert reporting, detection, containment, recovery or disclosure conduct visible, so no origin-x-break test can be answered for any lifecycle stage. The exploited vulnerabilities are attacker capability, not documented victim gaps (the case file itself codes one zero-day, for which the RM18 exclusion applies; the one one-month-old CVE exploited against the Middle Eastern financial firm is reported only as the actor's quick turnaround, with nothing said about that firm's patching). This is therefore an atypical, state-actor campaign case in the sense of the method spec's Unknown rule: reporting confirms targeting and impact only; entry vectors are partly documented but sit entirely on the attacker side and are not load-bearing for a victim-side mechanism. No stage distribution and no confidence score are issued.
Atypical state-actor campaign case: euRepoC records APT37's 2017 expansion of targeting to Japan, Vietnam and the Middle East as one case with diffuse, unnamed receivers (the case file carries no receiver_name; categories repeat three times for the three country groups), sourced only from FireEye's special report. DV-Type is Unknown per the decision rule (reporting confirms targeting/impact only; no victim-side mechanism is documented at any lifecycle stage), the stage and distribution keys are null and no confidence is issued; 0 risk markers are coded (Unknown allows 0-4) because no victim-side condition is evidenced - the zero-day (case file zero_days: Yes) and the quickly exploited recent CVEs are attacker capability, not documented victim gaps, and RM18's exclusion bars coding the zero-day. Victim established = false: the affected organisations (e.g. a Vietnamese international trading and transport company, a Middle Eastern financial company, an entity in Japan associated with UN sanctions/human-rights missions, per the FireEye report) are described but never named. Source caveat: the dataset's source_url https://www2.fireeye.com/rs/848-DID-242/images/rpt_APT37.pdf serves an HTML landing/bot-check page to non-browser clients, so the same FireEye report was read and cited from the Google-hosted copy (services.google.com); it is a specific document URL, not a homepage. Data quality: receiver_country lists Vietnam x9, Japan x9 and Middle East (region) x9 with no company names; weighted_intensity 3; the row's name field is an actor-vs-regions label, not a victim. Country left null: the method spec leaves it blank for multi-organisation campaigns spanning several countries (the checker's empty-Country warning refers to this); Sector: the method spec would also leave Sector blank for a diffuse multi-sector campaign, but the output check requires exactly one CBS SBI entry, so it is coded from the first-listed receiver category (government ministries / state institutions; defence industry also among the receivers) - researchers should review this forced choice. Year 2017 from eurepoc_extra.start_date 01.01.2017 (the year of the expansion); attribution and disclosure followed on 01.01.2018 via the FireEye report and blog. Unattended run: no consent-gated actions were required; no leak sites or attacker infrastructure were visited.