← Breach analysis by the Dual Vulnerabilities framework

Palestinian Authority staff FrozenCell mobile spyware surveillance (Gaza Cybergang / Two-tailed Scorpion, 2016)

2016 · Palestine · O - Public administration and defence · Mobile surveillanceware espionage campaign via fake apps and phishing

Analysed 2026-11-06 by Hermes

Open this case in the tool →

What happened

Mobile surveillanceware espionage campaign via fake apps and phishing

Dual vulnerability type

Expertise to Execution

Evidence coverage: 3/6 stages, weight 15/60

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

Why this classification

(1) The victims were people whose roles made their phones legitimate intelligence targets: employees of Palestinian government agencies, the security services, Palestinian students and Fatah-affiliated individuals; the expectation for state institutions whose staff handle sensitive material is that endpoints are hardened (sideloading and application control enforced, mobile malware protection deployed) and that published threat intelligence on an active local adversary is used. (2) In practice, a campaign that used no exploits at all - phishing and trojanised fake apps, including fake Facebook/WhatsApp/Messenger updates and a spoofed Tawjihi 2016 exam app - ran from at least 2016 (Unit 42 monitored KASPERAGENT from March 2016, with likely use from July 2015) until it was surfaced by outside researchers in 2017, exfiltrating 561MB from 24 Android devices (call recordings, SMS, geolocation, documents) and over 1GB in total, while the attackers remained active; the targeted organisations themselves never noticed. (3) The classification rests on the detection stage, where the dwell time and the external, vendor-driven discovery are best documented: the unactioned knowledge sat with technical specialists (endpoint hardening, app control, mobile threat detection and published reporting on this very actor were known, available remedies), and the chain broke in execution - those remedies were never realised in the victims' operations; no source documents a doctrinal artefact, a warning leaders ignored, or a leadership decision at the targeted organisations, so the gap is within the material column (Expertise to Execution). The entry vector (social engineering without exploits) is documented but not load-bearing for the DV-Type; the multi-year undetected surveillance is. Victim-side limits: the targets are a population across several Palestinian Authority institutions rather than one organisation, and no victim-side doctrine or response is documented anywhere, which is why the origin/break readings carry inference-level sources. (4) RM28 (Expertise to Execution lens): endpoint protection and application control were absent or ineffective across the targeted population - known sideloading and app-control safeguards were never deployed, so fake-update APKs installed and persisted on government-affiliated devices for years (causal role: Enabling). RM34 (Expertise to Execution lens): available detection capability was never configured, staffed or maintained at the victim side - the campaign ran from early 2016 to Lookout's October 2017 disclosure without any victim-side alert, even though Unit 42/ClearSky had already reported the group's toolset in April 2017 (causal role: Obscuring). RM35 (Expertise to Execution lens): thousands of hours of call recordings, SMS, images, geolocation and documents flowed continuously to attacker infrastructure (561MB found by Lookout from 24 devices, 1GB+ hosted) with no egress detection or prevention anywhere in the environment, although egress monitoring of such sensitive material should have existed (causal role: Amplifying).

Evidence by stage

Risk markers

Sources

Notes

The case file carries no receiver_name; the row's name field, Mobile malware FrozenCell, is the malware, not an organisation, and is kept as the Victim label as written in the case file. The description and Lookout's report identify the actual targets as a population spanning several organisations: employees of Palestinian government agencies, the security services, Palestinian students and Fatah-affiliated individuals. Sector set to O following the receiver categories (government/ministries, intelligence agencies, political parties; state institutions dominant). Attribution to Gaza Cybergang Group 2 / Two-tailed Scorpion (= APT-C-23 / Desert Falcons / Mantis; the case file's initiator_name adds Arid Viper, Grey Karkadann, UNC718, Renegade Jackal, Desertvarnish) rests on IT-security-vendor technical reporting; the Hamas link is an analyst suspicion reported by press, unconfirmed. All lifecycle evidence comes from security-vendor reporting (Lookout, published 05-10-2017 per the ORKL archive; Unit 42/ClearSky, April 2017) and secondary press (CyberScoop, 09-10-2017); no statement, notification or remediation by any Palestinian Authority organisation was found, which limits victim-side detail across recovery and response. Year 2016 from eurepoc_extra.start_date (01.01.2016), matching the earliest campaign evidence (Tawjihi 2016 lure; Unit 42 monitoring from March 2016). Exfiltrated data was encrypted by the attackers, but Lookout could decrypt it because key-derivation information sat in top-level directories - an attacker OpSec failure, recorded here only as context, not a victim-side marker. No related_rows; no earlier result for this victim in the output directory. Link check: blog.lookout.com blocks automated access (HTTP 403); the page was opened and read during research. No step required security-policy consent; no leak sites or stolen data were used.