← Breach analysis by the Dual Vulnerabilities framework

BBZ Schaffhausen Ransomware Attack (Unattributed, 2024)

2024 · Switzerland · P - Education · Ransomware attack via exploited firewall

Analysed 2026-11-06 by Hermes

Open this case in the tool →

What happened

Ransomware attack via exploited firewall

Dual vulnerability type

Expertise to Execution

Evidence coverage: 1/6 stages, weight 6/60

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

Why this classification

(1) The BBZ is the canton of Schaffhausen's vocational training centre (about 1,800 learners, about 170 staff); as a cantonal institution operating an internet-facing IT perimeter, the standing expectation was that its outward-facing firewall is kept remediated and hardened so that published product weaknesses do not become a gateway. (2) In practice, on 2 October 2024 unknown cybercriminals entered through a security gap in the firewall, ran encryption malware on the BBZ servers, blocked access to several systems and demanded a ransom; the school discovered it the next day, refused to pay, restored systems with internal and external experts, informed the police, the NCSC and the cantonal data protection officer as well as staff, parents, learners and training companies, and whether data had been stolen was still open at the public disclosure on 21 October 2024. (3) The classification rests on the entry stage, the only lifecycle stage with a documented victim-side gap: the unactioned expectation was material - remediating a weakness in perimeter firewall software is known technical-operational work - and the chain broke in execution, since the gap simply remained exploitable on the exposed appliance; no source documents any policy, standard, budget or leadership decision that was deferred, under-resourced or miscommunicated (so not Strategy to Leadership or Strategy to Execution), and no expert warning discounted by leaders is documented (so not Expertise to Leadership). The later stages are documented as aligned rather than gap-bearing: one-day detection, canton IT untouched, restoration running, authorities and affected parties notified, ransom refused - so the entry gap is not load-bearing through any post-breach stage. What the disclosure does not give is the specific CVE, the vendor, the patch timeline, or the eventual answer on data theft; the gap class (remediable firewall weakness) is nonetheless named by the victim itself. (4) RM18 (Expertise to Execution lens): the exploited security gap sat in the school's own firewall, a product whose weaknesses are remediable by vendor patches that the technical operators responsible for the appliance did not have realised on it when the attacker arrived, enabling the intrusion (causal role: Enabling). RM27 (Expertise to Execution lens): an internet-facing appliance was left in an unhardened, exploitable state at its exposure point - the disclosure's Einfallstor formulation marks exactly this external-exposure condition, whose technical configuration the school's operators did not maintain (causal role: Enabling).

Evidence by stage

Risk markers

Sources

Notes

Primary source is the Erziehungsdepartement Kanton Schaffhausen media release of 21.10.2024 (victim-state authority disclosure), which all opened reporting (inside-it.ch - the row's source_url - and The Record) restates. The BBZ's own Info_BBZ.pdf cited by aggregators no longer resolves (server returns an HTML page, not the PDF), so the canton's release on sh.ch is cited as the primary document. The victim's disclosure names only a security gap in the firewall - no CVE, vendor or patch timeline - so RM18 vs RM27 cannot be separated on the evidence and both facets of the perimeter failure are coded, with the origin/break answers marked inference for that reason; no zero-day indication exists, so RM18's exclusion does not fire. Detection (1 day), containment (canton IT unaffected), recovery (functional impact < 7 days per EuRepoC) and notification are documented as aligned, so no later-stage marker is coded; the 19-day gap between detection and public disclosure coincided with the scheduled autumn break and is not characterised as unreasonable by any source, hence no RM39. Dataset-quality remarks: eurepoc_extra.political_response_date 01.10.2024 predates the incident start of 02.10.2024 (no political response documented; the legal_response entry of 21.10.2024 naming the NCSC matches the disclosure/police report), and eurepoc_extra.end_date 03.10.2024 understates the recovery, which was still running at disclosure. Victim and country from receiver fields, year from start_date; no related_rows and no earlier incident at this victim found.