2016 · Poland · K - Financial and insurance activities · Malware infection via watering hole of the financial regulator's website, with data theft
Analysed 2026-11-06 by Hermes
Open this case in the tool →Malware infection via watering hole of the financial regulator's website, with data theft
Expertise to Execution
Evidence coverage: 5/6 stages, weight 27/60
Decisive stage: entry (clearest evidence) · Alternative: Strategy to Leadership
(1) The strategic expectation at the Polish banks was the supervisor-prescribed security framework itself: KNF required banks to raise security standards and keep their ICT infrastructure at an appropriate level (parliamentary interpellation), and within that framework the channel to the supervisor was trusted without verification - RCB notes that in the trust relationship it was simply assumed that KNF, which sets and enforces the standards, must itself be safe. (2) In practice, staff visits to the KNF information service - booby-trapped from 5 October 2016 to 2 February 2017 - delivered an IP-filtered exploit kit that exploited Silverlight and Flash vulnerabilities whose vendor fixes had been published between 28 December 2015 and 2 May 2016, installed a multi-stage RAT undetected by antivirus, and gave the attackers reconnaissance and control over key servers in some banks; encrypted data left the network unnoticed because traffic from infrastructure holding protected information was not monitored; the sector detected the intrusion only in late January 2017 (about four months' dwell) and disclosure came from the media on 3 February 2017. (3) This is a knowing-doing gap within the material column: every load-bearing remedy - browser plugin updates, endpoint detection, egress monitoring - was known, available technical expertise, and the chain broke in operations that never applied or implemented it; no source documents a bank leadership decision to defer patching or monitoring, so the break is execution, not leadership. The classification rests on the entry stage (clearest evidence, load-bearing); the egress gap points the same way. The structural condition that made the sector reachable sat with the supervisor: KNF kept its public information service in a vulnerable application environment, ignored warnings that the site contained poisoned code, and reported the interference attempt on 7 October 2016 yet the site stayed booby-trapped for four more months - read against KNF itself as victim, that ignored-warning pattern is a secondary Expertise to Leadership pattern, and the media-forced, classification-impeded disclosure shows a Strategy to Leadership tint at the response stage, but both are subordinate to the documented bank-side execution gaps. The entry vector is known and load-bearing. RM13 (Expertise to Execution lens): the banks' technical teams had no means to verify the integrity of the supervisor's website they depended on daily - the trust boundary was assumed without any assurance mechanism, which is how the attackers entered (causal role: Enabling). RM18 (Expertise to Execution lens): the exploited Silverlight and Flash vulnerabilities were known and remediable, with fixes published five to ten months before the October 2016 infections, yet bank workstations kept outdated vulnerable software (causal role: Enabling). RM35 (Expertise to Execution lens): egress monitoring over infrastructure holding protected information was not implemented at all, so encrypted reconnaissance data and files left continuously and even the victims could not later identify what was taken (causal role: Amplifying). RM37 (Expertise to Execution lens): the TLP:Amber classification of the sector IoC exchange prevented technical teams from sharing and operationalising evidence across institutions, so responders worked from fragments while scoping the intrusion (causal role: Constraining).
RM13 — Third-party / supply-chain trust boundaryRM18 — Vulnerability / patch latencyRM35 — Egress / exfiltration monitoring gapRM37 — Response coordination fragmentationCase file has no receiver_name; victim named from name/description and confirmed by BadCyber (at least a few of 20-plus commercial banks confirmed; BAE and Proofpoint also say at least 20 banks). Attribution is community-level (EuRepoC attribution_basis): BAE called the technical link to Lazarus unclear and Symantec initially had weak evidence, while Proofpoint states attribution to Lazarus/Bluenoroff is widely accepted across the industry; RCB treats Russian-language clues in one related sample as a probable false flag. The RCB article is a signed incident analysis by Polska Obywatelska Cyberobrona authors published on the government RCB archive site, quoting KNF's and the prosecutor's statements; treated as a strong secondary source, not a formal state report. The case file's multi-country receiver list (PL, US, MX, BR, DK, VE, CO, PE, UK, IN) mirrors the campaign's target spread (Symantec: 100+ organisations in 31 countries); Country is set to Poland as the primary victim of this row. Year 2016 from start_date 01.10.2016; RCB gives the KNF booby-trap window 5 October 2016 to 2 February 2017. Read against the supervisor rather than the banks, the ignored poisoning warnings at KNF form an Expertise to Leadership pattern (noted in the rationale). No action required consent: public sources only, no forms, logins or leak sites visited; unattended run.