← Breach analysis by the Dual Vulnerabilities framework

COMELEC voter data leak "Comeleak" (Anonymous Philippines / LulzSec Pilipinas, 2016)

2016 · Philippines · O - Public administration and defence · Website defacement and mass voter data theft

Analysed 2026-11-06 by Hermes

Open this case in the tool →

What happened

Website defacement and mass voter data theft

Dual vulnerability type

Strategy to Leadership

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

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

Why this classification

(1) COMELEC, the national election body holding biometric-quality voter registration data, carried an explicit doctrinal expectation: the Data Privacy Act of 2012 (RA 10173) obliged it as personal information controller to implement organisational and technical protections, designate a Data Protection Officer and manage breaches - all in force four years before the incident, around an election-critical system. (2) In practice, unknown actors extracted the voter, overseas-voter, firearms-ban and personnel databases from the public-facing web applications between 20 and 27 March 2016 (75.3m voter records, 15.8m fingerprint records, passport numbers in plain text) with no internal detection; COMELEC learned of it only when Anonymous Philippines defaced the site and LulzSec Pilipinas published the dump, and its officials publicly denied that sensitive data had been lost. The NPC's Case 16-001 decision found COMELEC violated Sections 11, 20 and 21 of RA 10173 and that at the time it had no data privacy policy, no Data Protection Officer, no privacy management program and no breach management procedure. (3) The gap is doctrinal-to-leadership, not a strategy failing in execution: the regulator placed the break at the decision level, on the agency head whose 'willful and intentional disregard of his duties as head of agency' was 'tantamount to gross negligence', noting that 'a head of agency making his acts depend on the recommendations of the Executive Director or the Information Technology Department' does not discharge the duty because 'the duty to obey the law should begin at the top'. The statutory expectation never reached an owner, a priority or a budget at leadership level, so no artefact was ever produced to operationalise it; the entry vector itself was never disclosed and is not load-bearing. The classification rests on the egress stage (clearest evidence, primary regulator decision); detection and response show the same doctrinal-leadership pattern. (4) RM16 (Strategy to Leadership lens): leadership commitments on data responsibility were absent from every decision and oversight mechanism - the NPC found no data governance policy at all, so crown-jewel fingerprint and passport data sat unprotected in public-facing lookup applications (causal role: Amplifying). RM34 (Strategy to Leadership lens): leadership oversight never made detection readiness real - a week-long extraction of the national voter register passed unnoticed and the agency learned of the breach from the attackers' announcement (causal role: Obscuring). RM35 (Strategy to Leadership lens): exfiltration and data-loss risk received no strategic attention - the entire database left the web tier over several days undetected and uninterrupted, and the NPC had to order security measures and a five-year audit cycle afterwards (causal role: Amplifying). RM39 (Strategy to Leadership lens): leadership disclosure behaviour diverged from legal expectations - the incident was disclosed only through the attacker's leak, and the first official account minimised the loss as containing 'no sensitive information', contradicted by the regulator's own inventory of the exposed data (causal role: Constraining).

Evidence by stage

Risk markers

Sources

Notes

Link check: newsinfo.inquirer.net blocks automated access (HTTP 403); the page was opened and read during research. Link-check: the case-file attribution URL (trendmicro.com research page) now serves only a site shell and the privacy.gov.ph decision article returns 404; the analysis cites the exact Wayback snapshots opened during this task (20220521165435 for Trend Micro, 20230309163928 for the NPC). The Guardian and BBC articles opened directly. Data quality: the row description says the leak covered 'fingerprints of more than 50 million persons'; Trend Micro and the NPC decision document 15.8 million fingerprint records within ~55 million active voter records (75.3m total records), so the fingerprint figure in the row is overstated. The row name says 'DDoS'; the opened sources document the 27 March defacement and the mass database theft - a distinct volumetric DDoS was not confirmed in the sources opened (the EuRepoC intensity field scores disruption as short-term, < 24h). Attribution of the two hacker groups rests on technical reports and the IT-security community (per the case file); the NPC additionally noted one computer used in the intrusion had an IP address registered to the National Bureau of Investigation and recommended a DOJ investigation - unresolved at the time of the decision. The NPC decision was taken into DOJ proceedings; Chairman Bautista disputed the gross-negligence finding (Inquirer). Context, not part of this row: NPC records show a second large-scale COMELEC breach in January 2017 (stolen election-office computer with voter/biometric data, compliance order February 2017), surfaced via search but not researched for this row.