← Breach analysis by the Dual Vulnerabilities framework

Evide ransomware / charity data theft (Unknown actors, 2023)

2023 · United Kingdom · J - Information and communication · Ransomware attack with data theft at a charity-sector data supplier

Analysed 2026-11-06 by Hermes

Open this case in the tool →

What happened

Ransomware attack with data theft at a charity-sector data supplier

Dual vulnerability type

Expertise to Execution

Evidence coverage: 2/6 stages, weight 12/60

Decisive stage: egress (clearest evidence) · Alternative: Strategy to Execution

Why this classification

(1) Evide acted as data processor for about 140 charities and community organisations, several of them serving rape and sexual-abuse survivors, running their service-user records on its Impact Tracker platform; the expectation resting with its technical specialists was that client datasets would be separated so one intrusion could not reach all of them, and that the platform's logging and egress controls would let it say what left the system. (2) In practice an unattributed ransomware intrusion in late March 2023 exfiltrated highly sensitive personal data (names, addresses, phone numbers, emails, short service-engagement records; about 2,000 people affected in Ireland alone), the data of all 140 client organisations was reportedly accessed in one tranche, and Evide could not reconstruct the exfiltration - it told customers to assume their data was stolen, and the nature of the stolen information remained officially unclear. (3) This is a knowing-doing gap within the material column: the remedies (tenant isolation, forensic/logging and egress capability) are standard technical practice that needs no leadership decision the record shows, yet they were not realised in the platform's architecture and operations; the entry vector is undisclosed and is not load-bearing, and the response itself (prompt police, NCSC and DPC notification, client and individual notification) was documented as functioning. The classification rests on the egress stage, corroborated by the containment stage; the origin/break answers are the coder's inference because no source documents Evide's internal controls. (4) RM30 (Expertise to Execution lens): the platform's architecture could not contain the compromise - the blast radius equalled the platform's entire footprint, with all roughly 140 client datasets accessed in one tranche, because the specialist containment remedy of per-client isolation was not implemented (causal role: Amplifying). RM33 (Expertise to Execution lens): investigation could not scope the incident after detection - Evide could not determine what data was taken and left customers to assume theft, showing the logging and forensic-readiness capability needed to reconstruct the exfiltration was missing in practice (causal role: Obscuring).

Evidence by stage

Risk markers

Sources

Notes

EuRepoC row; victim Evide (Derry/Londonderry, receiver_country United Kingdom in the case file), incident start 30.03.2023, disclosed by media and victim on 17.04.2023. Entry vector and dwell time undisclosed; actor unattributed (an unattributed actor does not lower the confidence). Ransomware was demanded but not paid and EuRepoC records no system disruption, so recovery is not in play as a constraint. Origin and break answers for the containment and egress stages are coder inferences - no source documents Evide's internal controls - hence those stages are typed as inference. Responsibility split controllers/processor (charities as data owners, Evide as processor, Irish DPC investigations of the charities possible per Irish Examiner) is noted but is a downstream-party issue, not coded as a marker for this victim. Egress monitoring beyond the forensic gap (no DLP evidence) is described in the rationale, not coded as RM35. RTÉ article content partly gated by a consent wall; article facts cross-checked against BBC, Belfast Live, Computer Weekly and Irish Examiner reporting. Link check: inishowen.ie blocks automated access (HTTP 403); the page was opened and read during research.