2022 · United States · O - Public administration and defence · Phishing intrusion and malicious use of legitimate RMM software in a refund scam
Analysed 2026-11-06 by Hermes
Open this case in the tool →Phishing intrusion and malicious use of legitimate RMM software in a refund scam
Strategy to Execution
Evidence coverage: 2/6 stages, weight 20/60
Decisive stage: detection (tie-break: post-breach stage) · Alternative: Expertise to Execution
(1) Strategic expectation: federal civilian organizations sit under a doctrinal cybersecurity posture - the CISA-operated, FCEB-wide EINSTEIN monitoring programme plus the joint CISA/NSA/MS-ISAC guidance that defenders audit remote-access tools, review logs for RMM software running as portable executables, apply application controls that allowlist authorized RMM programs and block portable versions, and block RMM ports at the perimeter. (2) In practice, employees at two FCEB organizations were deceived by help-desk-themed phishing in mid-June and mid-September 2022 into visiting malicious domains and running portable AnyDesk/ScreenConnect executables that ran in user context; CISA states this 'effectively bypass[es] common software controls and risk management assumptions' and that RMM use 'generally does not trigger antivirus or antimalware defenses'; neither organization noticed - CISA found the activity in October 2022 only by retrospectively analysing EINSTEIN data, and the campaign ran on into early November. (3) The classification rests on the detection stage, which wins the tie-break over the entry stage as the post-breach stage: the expectation was doctrinal (a federal programme and published guidance that name exactly these controls) and the break was in execution - the documented monitoring and application-control requirements were not realised in the organizations' operations, leaving months of undetected attacker access rather than merely weak security. The entry vector (phishing to portable RMM) is documented and shows endpoint-control ineffectiveness, but is not load-bearing for the DV-Type. (4) RM34 (Strategy to Execution lens): monitoring and alert-handling requirements were not operated as intended - the organizations' own detection never fired on weeks of bidirectional RMM traffic and the incident surfaced only through a retrospective federal-IDS review months after entry (causal role: Obscuring). RM28 (Strategy to Execution lens): endpoint and application-control requirements were not enforced across devices - unapproved portable RMM executables ran with local user rights despite installation audit/block controls, and antivirus/antimalware stayed silent on the legitimate tools (causal role: Enabling). A secondary Expertise to Execution pattern appears at entry (known remedies documented, not executed), but the decisive gap is the doctrinally-specified detection capability that was not realised.
RM34 — Monitoring / detection / alert-handling gapRM28 — Endpoint / malware-protection gapVictim organizations are unnamed: CISA's advisory identifies only 'two FCEB networks' (and later related activity on many other FCEB networks), so the victim is named descriptively; EuRepoC row carries no receiver_name. The financial damage fell on individual employees' personal bank accounts via the refund scam, not on the organizations - the victim-side DV evidence is the undetected access to federal networks. Two twitter.com links in source_url were not used as citations. Disclosure was by authorities of the victim state (CISA/NSA/MS-ISAC advisory, 25-01-2023); no organization-level statement exists, so the response stage could not be assessed. Actor unattributed; this does not lower confidence per the method spec.