2024 · Russia · O - Public administration and defence · Data theft via compromised employee email account
Analysed 2026-11-06 by Hermes
Open this case in the tool →Data theft via compromised employee email account
Strategy to Execution
Evidence coverage: 3/6 stages, weight 17/60
Decisive stage: egress (clearest evidence) · Alternative: Expertise to Execution
(1) STC is a sanctioned St Petersburg defence enterprise producing the Orlan-10 UAV and electronic-warfare systems and executing the Russian state defence order (InformNapalm CYBINT investigation; GUR announcement); the doctrinal expectation for such an organisation is that classified military documentation - drawings, specifications, patents, software - sits inside classified, access-controlled and monitored channels, separated from commodity data. (2) In practice, a procurement specialist of the R&D department handled and stored this documentation in his personal inbox.ru webmail account, which he used as a notepad; pro-Ukrainian hacktivists drew on that account as a source for roughly 10 months, and around 100GB - documentation for 194 nomenclature items, including projects where STC was only a co-executor - was handed to Ukraine's Defence Intelligence, announced on 8 January 2024; the company never detected anything itself and never acknowledged the incident publicly. (3) This is a knowing-doing / doctrinal-material gap, not merely weak security: the protection requirements for classified defence data exist at the strategic level by the very nature of the state-defence-order work (doctrinal origin), yet day-to-day operations ran the crown jewels through a consumer mailbox outside every company control, so no egress or monitoring capability could even see the data leaving (break at execution) - a formal expectation that never translated into operational reality. The entry vector (how the account fell into the attackers' hands, or how far the employee co-operated: reporting describes both a hacked account and a whistleblower who asked for help leaving Russia) is not documented and is not load-bearing for the classification; the classification rests on the egress / data-governance stage, with detection contributing. (4) RM16 (Strategy to Execution lens): classification and handling requirements for military documentation were not implemented in systems and processes - the technical archive of 194 equipment items, plus co-executed projects beyond STC's own remit, lived in one personal mailbox, so far more and more sensitive data was exposed than the work required (causal role: Amplifying). RM35 (Strategy to Execution lens): no data-transfer restriction or egress barrier operated on the channel that mattered - bulk classified material left the organisation's control over months without being detected or interrupted, and for a state-defence-order enterprise such visibility and prevention should have existed (causal role: Amplifying). RM09 (Strategy to Execution lens): personnel-security requirements were not implemented - a procurement insider was, by the attackers' own account, an important source for about 10 months, and insider monitoring neither prevented nor revealed it; the employee learned of the compromise only later and the organisation never did (causal role: Enabling). RM15 (Strategy to Execution lens): asset and data-flow visibility did not cover the exposure path that was actually used - classified work residing in an unmanaged personal webmail account was absent from any company inventory or map, which is why the ten-month collection was invisible (causal role: Enabling).
RM09 — Personnel security / insider gapRM15 — Asset / exposure visibility gapRM16 — Data governance / minimisation gapRM35 — Egress / exfiltration monitoring gapAttribution naming differs across sources: the EuRepoC row names One Fist (BBC attributes the January 2024 100GB weapons-maker hack to One Fist and shows its members receiving Ukrainian military certificates), while CBC and InformNapalm attribute the STC email exfiltration to the overlapping 'Cyber Resistance' label; both are treated as the same pro-Ukrainian hacktivist milieu in the Actor field. Mechanism evidence is largely attacker-friendly: the GUR announcement, InformNapalm (working with the hacktivists) and Defence Express describe what was taken and how it was obtained; the $1.5bn valuation is a GUR estimate with no stated derivation. Two acquisition narratives coexist: a hacked employee mailbox (CBC, InformNapalm, which also names procurement specialist Andrey Florinsky, who reportedly 'helped the hacktivists' and later complained he had been hacked) and a whistleblower who traded the data for help leaving Russia (Defence Express, 'Apostle pays money' initiative); the entry/detection-stage origin and break answers are read from these facts and labelled inference. STC itself never commented publicly (CBC: no response to a request for comment; no victim statement found), so victim-side governance evidence is indirect; per the search plan Russian-language searches surfaced no STC statement or regulator material. EuRepoC row's intelligence_impact wording (corruption without leaking) sits oddly with its own description of exfiltration; the description and all reporting describe pure data theft with no corruption or disruption, which is what was classified. Link check: informnapalm.org blocks automated access (HTTP 403); the page was opened and read during research.