← Breach analysis by the Dual Vulnerabilities framework

Equifax Data Breach (Chinese Military, 2017)

2017 · United States · J - Information and communication · Data theft via unpatched web-application vulnerability

Analysed 2026-11-07 by Hermes

Open this case in the tool →

What happened

Data theft via unpatched web-application vulnerability

Dual vulnerability type

Strategy to Execution

Evidence coverage: 5/6 stages, weight 45/60

Decisive stage: entry (tie-break: largest share) · Alternative: Expertise to Leadership

Why this classification

(1) Equifax is a national consumer reporting agency holding near-universal US adult identity data, under GLBA Safeguards-type duties and a formal internal doctrine: a 2016 Patch Management Policy with named owners, a certificate management process, deployed IDS/IPS (a Snort rule against this very Struts exploit installed 14 March 2017), and a legacy-modernisation effort begun before the breach. (2) In practice the doctrine was not realised: the US-CERT alert of 8 March 2017 was forwarded by Equifax's own GTVM team within a day with a 48-hour patching instruction, yet ACIS was never patched, both scanners missed the vulnerable installation, and attackers entered on 13 May 2017; the SSLV certificate feeding the IDS had been expired for 19 months, so the intrusion ran 76 days and the exfiltration was found only by chance at certificate renewal; unencrypted credentials on a file share let attackers pivot into 48 unrelated databases in a flat legacy environment described internally as having no segmentation between the Sun application servers and the rest of the network; roughly 145 million consumers' data left the network, and the announcement on 7 September 2017 found the company's website and call centres immediately overwhelmed. (3) The decisive stage is entry. The expectation that failed was doctrinal - a formal patch-management policy with a defined process and a 48-hour critical-patch instruction for this exact alert - and the chain broke at execution: the policy's roles had no official designees, scanning never located the system, and the patch was simply never applied. The House majority staff report frames this itself as the accountability gap producing an execution gap between IT policy development and operation, and the breach as entirely preventable: formal strategic artefacts stayed aspirational rather than operational, which is Strategy to Execution, not merely weak security. The containment gap shows a second pattern (Expertise to Leadership: a 2015 audit finding of missing segmentation that reached senior IT and security leaders and whose modernisation Equifax failed to prioritize), and the response gap a third (delayed disclosure with an unprepared consumer response), the entry vector is fully documented and load-bearing, but the classification rests on the victim's unexecuted patch policy rather than on the attacker's technique; among the two conflicting pre-breach conditions no source makes one cause the other, so the largest share decides, and entry (Strategy to Execution) beat containment (Expertise to Leadership) on the tie-break. (4) RM18 (Strategy to Execution lens): the vulnerability-management requirement - the 2016 policy and the 48-hour GTVM instruction - was not translated into timely scanning and patching: two scans missed the vulnerable Struts install and ACIS sat unpatched for two months on a bug the committee called entirely preventable (causal role: Enabling). RM30 (Strategy to Execution lens): segmentation and isolation requirements never reached architecture - with no segmentation between the internet-facing ACIS servers and the rest of the network, one web-application exploit yielded access to 48 unrelated databases and the crown-jewel identity tables (causal role: Amplifying). RM34 (Strategy to Execution lens): the monitoring requirement was deployed but not operated as intended - Snort rules and IDS/IPS stood behind an SSLV appliance whose certificate had expired for 19 months, file integrity monitoring was off, and the intrusion went undetected for 76 days (causal role: Obscuring). RM35 (Strategy to Execution lens): egress inspection existed on paper only - the encrypted-traffic appliance was dead, so 9,000 database queries and compressed PII files staged in a web-accessible directory left the network unimpeded and unnoticed (causal role: Amplifying).

Evidence by stage

Risk markers

Sources

Notes

Case-file lead ftc.gov/enforcement/refunds/equifax-data-breach-settlement returns 404 (and the FTC 2019 settlement press release 404s from this host), so the settlement and the FTC's allegations (the patch instruction was not followed up, sensitive data stored in plaintext) are cited from the opened BBC article; justice.gov serves a bot-challenge page, so the February 2020 DOJ indictment of four PLA 54th Research Institute officers is cited from the opened PBS/AP report. The case file's description (neglected cybersecurity; settlement) matches the stated victim. Related row 1042: no row-1042.json case file exists in the cases-run500 folder and no earlier result file exists; its dataset title describes a health insurance firm's database breach, which does not match Equifax, so no comparable earlier victim-side incident could be established. The six ACIS security concerns and the 2015 audit are recorded in the containment stage; a separate RM21 was not coded so the marker set stays on the four load-bearing stage mechanisms. Sector: consumer credit reporting is data-brokerage of consumer data (CBS SBI 63, section J).