2020 · Italy · J - Information and communication · Customer database theft sold on dark web
Analysed 2026-11-07 by Hermes
Open this case in the tool →Customer database theft sold on dark web
Strategy to Leadership
Evidence coverage: 2/6 stages, weight 14/60
Decisive stage: response (tie-break: post-breach stage) · Alternative: Expertise to Execution
(1) As an Italian mobile operator holding the anagrafic identity and SIM technical data of millions of subscribers, under GDPR notification duties and fresh AGCOM rules on SIM-change security, ho. Mobile was expected to detect a compromise of its customer database and to communicate accurately about public reports of a leak. (2) In practice the database was on sale on a dark web forum from December 22, 2020 without ho. Mobile noticing; when Bank Security made the sale public on December 28, the company told the press it had no evidence of massive illegal access to its systems, and only on January 4, 2021 - after researchers had verified sample records with affected subscribers - did its own official statement confirm that anagrafic and SIM technical data of part of its customer base had been stolen, while entry and containment were never explained in any source. (3) The classification rests on the response stage: the expectation that was not met was doctrinal (the organisation's official public communication), and the chain broke at leadership, because the public denial was contradicted within a week by the company's own disclosure - leadership communicated the situation inaccurately while customers stayed exposed and a buyer may already have purchased the database. The detection gap (external discovery of a sale running for six days) points to Expertise to Execution, but no source documents its origin and break, so as an inference it cannot win the tie-break against the documented response stage. (4) RM34 (Strategy to Leadership lens): leadership oversight did not make detection readiness effective - the intrusion and the bulk exfiltration surfaced only through an outside analyst's dark-web monitoring, and the company's own December 28 position was that it had no evidence of any illegal access (causal role: Obscuring). RM39 (Strategy to Leadership lens): leadership disclosure behaviour diverged from what its own subsequent investigation showed - a public statement ruling out a compromise on December 28/29, reversed by the January 4 admission of theft, with customer notification by SMS only on January 4 (causal role: Constraining).
RM34 — Monitoring / detection / alert-handling gapRM39 — Breach notification / disclosure failureinformationisbeautiful row: victim and one-line description only; all lifecycle evidence from own research. The source_url in the case file carries a trailing ' /' (data-quality defect) and the linked ZDNet article matches the victim. 2.5M records is the researcher's (Bank Security) estimate; ho. Mobile's own statement only says 'part of the customer base' - the size is therefore unconfirmed by the victim. Entry vector was never disclosed in any source and is not load-bearing. Positive response measures after January 4 (free SIM replacement with in-person identity checks, automatic ICCID regeneration on January 9, Garante and police notification) are documented but do not offset the initial inaccurate public denial. A Security Report note (December 30, 2020) that an API returning ho. Mobile user data may have existed since 2019 is an unverified forum claim and was not coded. No Garante privacy fine for this incident was found in the sources opened.