← Breach analysis by the Dual Vulnerabilities framework

Uniprévoyance Ransomware (Unattributed, 2024)

2024 · France · K - Financial and insurance activities · Ransomware attack with personal data theft

Analysed 2026-11-06 by Hermes

Open this case in the tool →

What happened

Ransomware attack with personal data theft

Dual vulnerability type

Strategy to Execution

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

Decisive stage: recovery (clearest evidence) · Alternative: Expertise to Execution

Why this classification

(1) Uniprévoyance, a paritary provident institution protecting around 1.5 million people and supervised by the ACPR, held formal strategic artefacts for cyber continuity: it could and did trigger its business-continuity plan (PCA) and disaster-recovery plan (PRA) the same evening, and by 2024 it was preparing DORA compliance - the doctrinal expectation of tested resilience and recoverability clearly existed. (2) In practice the artefacts did not translate into operational capability: the ransomware deployed on 20 May 2024, the institution stayed under its continuity and recovery plans until 12 July 2024 (53 days), the annual report records significant operational consequences including projects postponed to 2025, and the remediation programme for 2024-2025 consisted of restoring the systems with an accelerated investment programme for IT security, a DORA gap analysis and action plan, the definition of a new PRA and implementation of security-audit recommendations. Personal data including RIB bank details and claims data were compromised, and at disclosure the institution could not yet establish which data had left its systems. (3) The decisive stage is recovery, the stage with the clearest, primary documentation; the doctrinal column was populated (PCA, PRA, DORA programme) yet the material recovery reality diverged from it (multi-week restoration, a PRA that had to be rewritten, a security posture whose strengthening needed to be accelerated after the fact), which is the Strategy to Execution diagonal rather than mere weak security or a leadership refusal - no source shows leadership knew and deferred. Detection and containment were prompt and effective (in-house detection at 18:30, isolation the same evening, no propagation to partners), the response was compliant and transparent, and the entry vector is undisclosed and not load-bearing. RM31 (Strategy to Execution lens): the continuity requirement existed as a doctrinal artefact but was not translated into an effective, tested recovery capability - 53 days under the PRA/PRA-driven resumption and a new PRA written only afterwards (causal role: Constraining). RM33 (Strategy to Execution lens): logging and evidence-preservation requirements were not implemented well enough to scope the incident - the victim's own disclosure says investigations showed confidentiality was compromised while it could not yet determine which data were out (causal role: Obscuring). RM35 (Strategy to Execution lens): egress controls and data-transfer restrictions were not translated into effective operational barriers - banking (RIB) and claims data left an insurer that holds them under supervisory data-protection duties without prevention or timely detection (causal role: Amplifying).

Evidence by stage

Risk markers

Sources

Notes

Source caveats: the case-file source article (L'Argus de l'assurance, 14 June 2024) is subscriber-only - the original URL returns HTTP 403, only the freely visible summary was captured from it; the article was verified via the opened Internet Archive snapshot cited instead. A follow-up L'Argus piece (19 June 2024, investigation closed, no propagation to partners, health activities unaffected, restoration due early July) is blocked by a JS wall and no archive snapshot was reachable (Internet Archive was intermittently offline during this run); its content was taken from search-result summaries and is corroborated by the victim's annual report, so it is cited only in the containment stage, not in a scored claim. The dataset records no functional disruption, which understates the victim's own account of 53 days under PCA/PRA with projects postponed - the victim's disclosure was preferred. The entry vector is undisclosed and the attack unattributed; no public leak-site claim tied to this incident was used (policy excludes leak sites). As of 06-11-26 no CNIL sanction or ACPR/ANSSI enforcement against Uniprévoyance appears in the CNIL's public sanctions listings. The February 2024 Viamedis/Almerys third-party payment-processor incident that also affected several insurers is a separate incident and is not conflated here. An aggregator entry (csidb.net) mixes that separate incident into this case's record; it was not used as evidence.