← Breach analysis by the Dual Vulnerabilities framework

Central Hudson Customer Data Breach (Unattributed, 2013)

2013 · United States · D - Electricity, gas, steam and air conditioning supply · Data theft via malware (worm) intrusion into payment-processing systems

Analysed 2026-11-06 by Hermes

Open this case in the tool →

What happened

Data theft via malware (worm) intrusion into payment-processing systems

Dual vulnerability type

Strategy to Execution

Evidence coverage: 3/6 stages, weight 25/60

Decisive stage: detection (tie-break: pre-breach condition) · Alternative: Expertise to Execution

Why this classification

(1) Central Hudson is a regulated New York electric utility holding bank-account and check data for about 294,000 customers; the PSC order states that the basic controls Dell SecureWorks recommended (frequent antivirus updates, rigorous password-policy enforcement, inventory and destruction of stored customer information) were already part of its cybersecurity programme, so the doctrinal expectation existed on paper. (2) In practice a self-updating worm that disables antivirus and malware detection sat on the business network for at least six months before discovery on 16-02-2013, spread internally, secured persistent remote access, reached the cash-and-check payment processing system, and made self-initiated outbound file transfers of which neither the quantity nor the content could ever be determined, forcing the utility to treat the PII of approximately 110,000 customers as compromised. (3) The classification rests on the detection stage: the expectation originated doctrinally (the failed controls were programme content, per the PSC's own note) and the chain broke in execution (the programme's endpoint and monitoring requirements were not applied frequently or rigorously enough for the disabled antivirus and the six-month dwell to be noticed) - formal strategic artefacts that do not translate into operational reality, which is Strategy to Execution. The entry vector was never identified by forensics and is not load-bearing. The egress gap is a secondary Expertise to Execution pattern (DLP and traffic restriction, specialist-recommended remedies, absent in architecture); it lost the tie-break to the pre-breach detection condition. Response was the aligned counter-example: prompt regulator and customer notification and remediation. RM34 (Strategy to Execution lens): the programme's monitoring and antivirus-update requirements were not operated as intended - the worm was present at least six months and disabled antivirus on a desktop without triggering notice, so detection came only from chance observation of anomalous logon attempts (causal role: Obscuring). RM28 (Strategy to Execution lens): endpoint malware-protection requirements existed in the programme but were not enforced across devices - the malware disabled antivirus software and then spread over the network to the payment system, where application whitelisting was only a post-incident recommendation (causal role: Enabling). RM35 (Strategy to Execution lens): egress controls and DLP were never translated into operational barriers, so the worm's self-initiated outbound transfers were undetected and the loss could not even be scoped (causal role: Amplifying). RM16 (Strategy to Execution lens): retention and minimisation requirements were not operationalised in the systems - scans and records of customer checks remained stored on the infected payment-processing system, and Dell had to recommend more frequent inventory and destruction of stored customer information (causal role: Amplifying).

Evidence by stage

Risk markers

Sources

Notes

informationisbeautiful row: one sentence plus a source_url that no longer resolves to the article (the esecurityplanet link now redirects to a generic Threats archive page); the cited page was replaced with the New York Public Service Commission order in Case 13-M-0178 (19-08-2013), a primary source containing the Dell SecureWorks findings, and contemporaneous Daily Freeman reporting (20-02-2013). Incident: discovered 16-02-2013, disclosed 19-20-02-2013, ~110,000 of 294,000 customers potentially affected; the row's description matches the victim. Actor unattributed; entry vector never determined by forensics - deliberately not load-bearing. No related_rows in the case file. The doctrinal origin of the detection gap rests on the PSC order's footnote that the recommended basic measures were already part of the cybersecurity programme.