← Breach analysis by the Dual Vulnerabilities framework

Sanrio SanrioTown database exposure (Unattributed, 2015)

2015 · Japan · N - Administrative and support service activities · Data exposure via publicly accessible database (no intrusion)

Analysed 2026-11-06 by Hermes

Open this case in the tool →

What happened

Data exposure via publicly accessible database (no intrusion)

Dual vulnerability type

Expertise to Execution

Evidence coverage: 4/6 stages, weight 24/60

Decisive stage: entry (tie-break: largest share) · Alternative: Strategy to Execution

Why this classification

(1) A company running a child-facing online game and fan community that held members' personal data carried the operational expectation that its member databases were unreachable from the open internet and password records protected to prevailing practice, plus its own stated rule of no accounts for children under 13 and Hong Kong's DPP4 data-security duty. (2) In practice a misconfigured MongoDB installation - the company's own 'server misconfiguration' - left the sanriotown.com database and two mirrored backups readable online for about four weeks (earliest logged 22 November 2015), findable by anyone with a browser via Shodan and discovered by an external researcher on 19 December rather than by any internal monitoring; the exposed data included names, reversible birthdates, 186,261 under-18 accounts, unsalted SHA-1 hashes and plaintext password-hint answers; the advisory then called the unsalted SHA-1 'securely encrypted'. (3) This is a knowing-doing gap inside the material column, not merely weak security: the remedies - not exposing a database, scanning one's own external footprint, salting password hashes - were standard technical practice that needed no leadership decision or missing budget, yet were not executed; no source documents any policy or decision behind the exposure, so the entry and detection gaps are material-origin, execution-break. The egress gap is doctrinal-origin, execution-break (the company's own age rule versus the minors it actually held), a Strategy to Execution pattern, and the response gap (public security claim contradicted by the documented hashing) a Strategy to Leadership pattern. Two load-bearing pre-breach conditions conflict (entry/detection point to Expertise to Execution, egress to Strategy to Execution), no source shows either caused the other, and the response gap is post-breach and only contributing, so by the largest-share tie-break among parallel pre-breach conditions the classification rests on the entry stage and the case is Expertise to Execution; the entry vector here is not an attack at all - the exposure is the breach. RM27 (Expertise to Execution lens): operations could not keep a secure configuration against the organisation's own external footprint - a MongoDB installation left open and discoverable until an outsider found it, a known hardening remedy never executed (causal role: Enabling). RM34 (Expertise to Execution lens): the monitoring and exposure-scanning means available to a site like this were not configured or maintained to see its own reachable servers, giving an almost four-week dwell time closed only by an external party (causal role: Obscuring). RM16 (Expertise to Execution lens): data-governance obligations were never operationalised - the no-under-13 rule, minimisation and classification of member data existed on paper only, so 186,261 minors' records sat in one unhardened community database that also fed several other Hello Kitty portals (causal role: Amplifying). RM26 (Expertise to Execution lens): prevailing password-protection practice (salted, slow hashing; no plaintext recovery hints) was a known remedy that was not implemented - unsalted SHA-1 hashes and plaintext hint answers turned the exposure into crackable credentials (causal role: Amplifying).

Evidence by stage

Risk markers

Sources

Notes

The row's description matches the reporting. The breached site was operated by Sanrio Digital (HK) Limited, a joint venture (Typhoon Games 70% / Sanrio 30%), not by Sanrio Co. directly (AP); the dataset labels the victim Sanrio and the result follows it. Country: Sanrio Co. is Japanese (victim HQ per column convention); the operating company and the regulator (PCPD) are Hong Kong. Sector: Sanrio's primary activity is creation and licensing of characters (SBI 77.4, section N); the operating JV (games/community site) would fall under J. The Hong Kong PCPD opened an investigation on 23-12-2015; no published outcome could be found, so no regulator finding is available for or against alignment. Cause detail: Sanrio said it was 'possible (but not yet certain) that maintenance conducted on November 20th resulted in the database becoming accessible' (Salted Hash) - too uncertain to code RM20 change control. Salted Hash noted an earlier database leak of ~6,000 Sanrio shareholders' records earlier in 2015; not coded as RM21 because no source shows the two leaks shared an unremediated weakness. No malicious intrusion: the only confirmed access was the reporting researcher's; Sanrio's claim that no data was stolen rests on its own investigation. Unattributed actor does not lower confidence. Sector/country choices are judgement calls, explained above; incident_id is not carried by this row. Link check: databreaches.net blocks automated access (HTTP 403); the page was opened and read during research.