← Breach analysis by the Dual Vulnerabilities framework

Ukrainian CEC DDoS attacks (Suspected Russian state actors, 2019)

2019 · Ukraine · O - Public administration and defence · DDoS attack against election infrastructure

Analysed 2026-11-06 by Hermes

Open this case in the tool →

What happened

DDoS attack against election infrastructure

Dual vulnerability type

Expertise to Leadership

Evidence coverage: 1/6 stages, weight 8/60

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

Why this classification

(1) The Central Election Commission runs Ukraine's nationally critical election information infrastructure, and keeping its information resources reliable and uninterrupted was an explicit strategic commitment - the Verkhovna Rada passed a dedicated cyber-defense funding act (draft law No. 8496) for the CEC in November 2018. (2) In practice, on 24-25 February 2019, weeks before the presidential election, a two-day HTTP-flood DDoS from a botnet of non-updated WordPress sites blocked user access to the CEC's web resources (the EuRepoC dataset rates the disruption at over 24 hours; no data was stolen), and the Ukrainian authorities - while presenting the attacks as repulsed - acknowledged negative impacts that specialists had to stop. (3) The victim's own specialists had flagged the condition long before: the CEC IT Department chief said the work was hindered by a lack of financing, and the CEC chairman's stated need of at least Hr 36 million to secure the election infrastructure had not been disbursed less than a year before the vote; the funding decision came late (November 2018) and was not realised as effective availability protection before the attack. Experts warned and the decision level did not resource in time: the origin of the expectation is material (the known remedy and the constraint reports sit with technical specialists) and the break is at the leadership/decision level (funding deferred, not prioritised in time) - Expertise to Leadership. The entry vector is fully disclosed (HTTP flood) but is the attacker's technique and not load-bearing; the classification rests on the entry/exposure stage - the under-resourced availability protection - which is the only stage with a documented victim-side gap. A secondary doctrinal reading (the strategic artefact of November 2018 not yet materialised in operations - Strategy to Execution) is visible but weaker, since no source shows an approved, resourced programme failing in execution. (4) RM08 (Expertise to Leadership lens): the CEC's specialists reported their resourcing constraint publicly before the election - the IT chief citing a lack of financing and the chairman's Hr 36 million requirement left undisbursed - and leaders under-prioritised the constraint they were told about (causal role: Enabling). RM32 (Expertise to Leadership lens): DDoS-protection capacity was insufficient to keep the CEC's election web resources available under a two-day HTTP flood, and the capacity investment had been declined/deferred despite these expert warnings (causal role: Enabling). RM21 (Expertise to Leadership lens): the 2014 compromise of the same victim's election systems - the system found non-functional three days before the 2014 vote and a fake results page blocked on election night - was documented, yet its availability lesson had not been turned into funded, in-place protection by February 2019 (causal role: Perpetuating).

Evidence by stage

Risk markers

Sources

Notes

Attribution caveat: the case file carries initiator Unknown (category State); attribution to the Russian Federation is a direct statement by the victim-state president (26.02.2019) while the SBU said on 27.02.2019 it was still verifying possible involvement of Russian intelligence services - the DV is assessed on the victim side and the unconfirmed actor does not lower the score. Impact caveat: the long-term (>24h) disruption rating comes from the EuRepoC record itself (start 24.02.2019 / end 25.02.2019); the opened official statements present the attacks as thwarted/repulsed, and no independent reporting was found that fixes the exact website downtime - the incident impact is therefore taken from the dataset record while the mechanism rests on the opened sources. The victim's 2014 election-system compromise (system down three days before the vote, fake Yarosh results page) is documented in the Kyiv Post and Atlantic Council sources; it is a separate, earlier incident at the same victim and is used only as the comparable precedent for RM21. The Kyiv Post archive page and the ukrinform article about the bill were read via full-text extraction (curl + trafilatura) because page extraction otherwise returned navigation only. Research used public news, agency and think-tank sources only; no leak sites or criminal forums were consulted; nothing in this task required consent actions.