← Breach analysis by the Dual Vulnerabilities framework

Sega Pass data theft (unattributed, 2011)

2011 · Japan · J - Information and communication · Theft of customer database records

Analysed 2026-11-06 by Hermes

Open this case in the tool →

What happened

Theft of customer database records

Dual vulnerability type

Strategy to Execution

Evidence coverage: 2/6 stages, weight 14/60

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

Why this classification

(1) Strategic expectation: after the April 2011 PlayStation Network breach, Sega explicitly made protecting its online customer base a strategic priority - it brought its periodic security audit forward by about six months, made changes to its security systems in direct response to the attack wave on gaming companies, and publicly claimed certified privacy and storage of data on secure servers or computers inaccessible by modem. (2) In practice: about a month later, an unauthorised entry to the Sega Pass database succeeded; Sega confirmed that the names, birth dates, e-mail addresses and encoded passwords of all 1,290,755 customers had been brought outside of the system, and the bulk export was not detected or interrupted - the company caught the access, not the exfiltration. The intrusion vector and dwell time were never disclosed, so the entry technique is not load-bearing here. (3) The classification rests on the exposure/entry stage (clearest evidence): the expectation that failed was doctrinal in origin - Sega's own security review programme and post-PSN hardening decisions - and the chain broke in execution, because GamesIndustry.biz reports the changes made in the wake of the Sony breach failed to protect customer data and that the breach raises questions over the scope and implementation of Sega's internal security review: formal strategic artefacts existed and were acted on at decision level but did not translate into the operational protection of the live customer database. It is not Strategy to Leadership, because no source shows leadership deferring or under-resourcing a known need; not Expertise to Execution, because no specific technical remedy known to specialists and left unapplied is documented. The post-breach response was largely aligned (service down, all users e-mailed, passwords reset within a day of detection), so no post-breach stage competes for decisiveness; the full-table, undetected exfiltration points to the same doctrinal-to-execution pattern at the egress stage. (4) RM12 (Strategy to Execution lens): the assurance Sega relied on was disconnected from deployed exposure - the audit was a good housekeeping exercise that left the company pretty solid in its CEO's words, yet the whole customer database was stolen within a month, so the review materially misrepresented control effectiveness (causal role: Obscuring). RM35 (Strategy to Execution lens): the publicly stated secure-server and privacy-certification commitments were not translated into effective operational barriers on data leaving the network - the entire 1,290,755-record customer table was exported without detection or interruption (causal role: Amplifying).

Evidence by stage

Risk markers

Sources

Notes

Source caveats: the row's ZDNet source_url could not be opened (the blog post is gone); the cited articles were opened instead. The CNN article's live URL is dead, so its 23-06-2011 web.archive.org snapshot (opened) is cited. Entry vector and dwell time were never disclosed - Dark Reading: Sega did not say how the database was breached; the entry technique is therefore not load-bearing. Actor unattributed: LulzSec publicly denied involvement and offered to help find the attackers; no attribution followed. Disclosure caveats: English-language users saw only a maintenance message while the detailed company statement appeared in Japanese on the Japanese site (CNN), but registered users were e-mailed directly within roughly a day of detection, so no notification failure is coded; the public ESRB Privacy Certified claim is described in the rationale rather than coded as RM44 because no source establishes the claim's state before the breach. An outside consultant publicly doubted that the encoded passwords were safe against offline cracking (Dark Reading); no evidence on hashing quality exists, so RM26 is not coded. No regulator or formal investigation of the incident was found in reporting; the egress-monitoring gap is an inference from the victim's own statement of a complete, undetected export. The dataset description matches the stated victim (no data-quality mismatch). incident_id is not carried by this informationisbeautiful row.