Skip to navigation

Threat reportAPTTL-2026-5962

SilentAir: New State-Level Actor Intercepting Calls and SMS via SS7/Diameter Signaling from Hijacked Mobile Operator Nodes

highACTIVE

Enea reports SilentAir, a previously unknown state-level threat actor (high confidence) that launches mostly SS7 and occasionally Diameter signaling attacks from hijacked mobile operator network nodes to redirect targets' incoming calls and SMS and harvest subscriber data. Taiwan is the primary target; activity has been monitored since early 2025 and is assessed with low confidence as China-nexus.

Key facts

IdentifiersTL-2026-5962
SeverityHIGH
First disclosed
Affected productsVarious mobile network operators SS7 / Diameter signaling nodes and core-network hosts
AttributionSilentAir (low confidence)
Nation-state nexusChina
MalwareGTPDOOR, SCTPE, SilentAir
ToolsCordscan, SIGTRANslator
MITRE ATT&CK techniquesT1014 Rootkit, T1021.004 SSH, T1036.005 Match Legitimate Resource Name or Location, T1046 Network Service Discovery, T1070.003 Clear Command History, T1070.006 Timestomp, T1071.004 DNS, T1078 Valid Accounts, T1090 Proxy, T1095 Non-Application Layer Protocol, T1110.001 Password Guessing, T1111 Multi-Factor Authentication Interception and 8 more in the attack flow
MotivationESPIONAGE
Target sectorsgovernment administration, telecoms, political
Target regionstaiwan, Southeast Asia, East Asia, South Asia, 143 - Central Asia, East Africa, Middle East, Europe
Detection rules9 (1 shown free)
Sources7 cited sources

Detection coverage

As of 2026-11-13, Threadlinqs Intelligence publishes 9 detection rule(s) for TL-2026-5962 across Splunk SPL, Microsoft KQL and Sigma, covering 27 indicator(s) of compromise. The whole corpus is readable without an account. Every threat comes with one complete detection in Splunk SPL, Microsoft KQL and Sigma, free for everyone; a free account adds daily TL Creds to unlock the rest. Paid plans add a larger daily allowance, correlation and the MCP server. Threadlinqs MCP server · View plans.

Free detection: SilentAir SS7/Diameter Interception — UpdateLocation Followed by Silent SMS (Exploitation)

Behaviour-first detection of the SilentAir interception pattern reported by Enea (TL-2026-5962): an SS7 UpdateLocation or Diameter Update-Location-Request that redirects a subscriber's voice/SMS, followed within an hour by silent SMS (SMS-PID 64 / silent type) to the same subscriber to verify interception is in place. Enea also observed 2FA codes arriving immediately after interception. Covers compromised operator-node abuse (T1584) and adversary-in-the-middle interception (T1557.001). Requires normalized signaling-firewall logs.

  • Languages: Splunk SPL, Microsoft KQL, Sigma
  • ATT&CK: T1584, T1557.001
  • Data sources: telecom_signaling:signaling:firewall, SignalingFirewall_CL, ss7_signaling_firewall
  • Severity: critical
  • Confidence: high

Splunk SPL

index=telecom_signaling sourcetype="signaling:firewall" earliest=-24h
```SilentAir pattern: update-location interception followed by silent SMS verification for the same subscriber```
| eval op=lower(signaling_op)
| eval stage=case(match(op,"^(updatelocation|update-location-request|ulr)$"),"intercept", match(op,"^(mt-forwardsm|forwardsm|mt-forward-short-message)$") AND (tostring(sms_pid)="64" OR like(lower(sms_type),"%silent%")),"silent_sms", true(),null())
| where isnotnull(stage)
| eval sub=coalesce(imsi,msisdn)
| stats min(eval(if(stage="intercept",_time,null()))) AS first_intercept max(eval(if(stage="silent_sms",_time,null()))) AS last_silent count(eval(stage="silent_sms")) AS silent_sms_count dc(stage) AS stages values(src_gt) AS src_gt values(smsc_gt) AS smsc_gt BY sub
| where stages=2 AND last_silent>=first_intercept AND (last_silent-first_intercept)<=3600
| eval first_intercept=strftime(first_intercept,"%Y-%m-%d %H:%M:%S"), last_silent=strftime(last_silent,"%Y-%m-%d %H:%M:%S")
| table sub first_intercept last_silent silent_sms_count src_gt smsc_gt

Microsoft KQL

let Window = 1h;
let SilentSms = SignalingFirewall_CL
| where TimeGenerated > ago(24h)
| where tolower(Operation) in ("mt-forwardsm", "forwardsm", "mt-forward-short-message")
| where tostring(SmsPid) == "64" or tolower(SmsType) has "silent"
| project SilentTime = TimeGenerated, Imsi, SmscGt;
SignalingFirewall_CL
| where TimeGenerated > ago(24h)
| where tolower(Operation) in ("updatelocation", "update-location-request", "ulr")
| project UlTime = TimeGenerated, Imsi, SrcGt
| join kind=inner SilentSms on Imsi
| where SilentTime between (UlTime .. (UlTime + Window))
| summarize FirstUpdateLocation = min(UlTime), SilentSmsCount = count(), SrcGts = make_set(SrcGt, 10), SmscGts = make_set(SmscGt, 10) by Imsi
| order by SilentSmsCount desc

Sigma

title: SilentAir-Style Update Location From Untrusted Global Title
id: 3f6b1c52-8a4e-4d17-9b0a-5c2e7d41a601
status: experimental
description: Detects SS7 UpdateLocation or Diameter Update-Location-Request received from a Global Title or Diameter peer not on the trusted-partner list, the interception primitive used by SilentAir to redirect voice and SMS. Correlation with follow-up silent SMS is performed in the SPL and KQL flavors.
references:
    - https://www.enea.com/insights/the-enemy-within-tracking-liminal-panda-and-silentair-through-mobile-networks/
    - https://www.securityweek.com/silentair-hackers-intercept-calls-sms-from-inside-mobile-networks/
author: Threadlinqs Team
date: 2026-11-13
tags:
    - attack.resource-development
    - attack.t1584
    - attack.credential-access
    - attack.t1557.001
logsource:
    product: ss7_signaling_firewall
    service: events
detection:
    selection:
        Operation|contains:
            - 'UpdateLocation'
            - 'Update-Location-Request'
        SrcGtTrusted: false
    condition: selection
falsepositives:
    - Roaming partner not yet onboarded in the trusted GT list
    - Steering-of-roaming platforms sending legitimate location updates
level: high

Known false positives:

  • Legitimate roaming UpdateLocation followed by a network-originated silent SMS used for reachability checks by the home operator
  • Steering-of-roaming platforms that send location updates and test SMS for the same IMSI

Tuning notes: Map signaling_op, imsi, msisdn, src_gt, smsc_gt, sms_pid and sms_type to your signaling firewall's field names (these are normalized names, not a specific vendor schema). Exclude trusted roaming-partner GTs and your own SMSC GTs via a lookup, and widen or tighten the 3600-second correlation window. Prioritise results for VIP / government subscribers.

8 more detection rules (Splunk SPL, Microsoft KQL, Sigma) · unlock with TL Creds. Sign up free to unlock the detection rules or query them with the Threadlinqs MCP server (Purple)
  • SilentAir Signaling Probe-to-Interception Chain From a Single Global Title (Exploitation) · unlock with TL Creds
  • Anti-Forensics — Timestomping and History Clearing on Telecom Linux Hosts (Post-Exploitation) · unlock with TL Creds
  • NoDepDNS DNS C2 Channel and DNS-Tunnel Volume From Core Hosts (C2) · unlock with TL Creds
  • SilentAir Silent-SMS Burst Against a Single Subscriber (Exploitation) · unlock with TL Creds
  • CL-STA-0969 / LIMINAL PANDA Tooling, Hashes and Service Masquerading on Linux Core Hosts (Post-Exploitation) · unlock with TL Creds
  • SSH Password Guessing Then Valid-Account Login on Core-Network Linux Hosts (Lateral Movement) · unlock with TL Creds
  • PAM Backdoor — pam_unix.so Tampering on Telecom Core Hosts (Post-Exploitation) · unlock with TL Creds
  • GTPDoor Covert C2 Over GTP-C [indicator] Across IPX/GRX (C2) · unlock with TL Creds

Indicators of compromise

As of 2026-11-13, Threadlinqs Intelligence tracks 27 indicators of compromise for TL-2026-5962.

Sample indicators (3 of 27)

  • Domain: pgw-s5s8[.]mpgw001[.]node
  • File hash: 35f0d4d0c5519a2957338e448fce53bc
  • File hash: da35d98c9a22e8ede93595d21fe7e63e

Defanged, TLP:CLEAR. Full set of 27 indicators: unlock with TL Creds.

27 indicators of compromise · unlock with TL Creds. Compare plans

Community OSINT corroboration

1 of this threat's indicators have also been reported by the open-source security community. Community sightings are unverified and are kept separate from Threadlinqs' curated indicators. Indicator values, reporters and campaign linkage are available to authenticated Red-tier users.

Technical analysis

SilentAir is a signaling-layer threat actor first publicly documented by Enea (blog 'The Enemy Within: Tracking LIMINAL PANDA and SilentAir Through Mobile Networks', published 11 Nov 2026, updated 12 Nov 2026) and reported by SecurityWeek on 13 Nov 2026. Enea assesses with high confidence that SilentAir is a state-level actor rather than a commercial surveillance company, and with low confidence that it aligns with China-nexus operations, citing the primary focus on Taiwan, the absence of Chinese networks or subscribers among targets, and activity timing that matches a UTC+8 working day (starting around 09:00, dips in morning and evening, near-absent on weekends, pauses around Chinese holidays).

The actor sends signaling traffic from inside compromised network nodes belonging to mobile operators in at least six Asian countries (others suspected); Enea confirmed the source is a compromised node rather than a spoofed external origin. It uses SCCP Global Titles that resemble legitimate network elements and, unlike the surveillance-vendor actor HiddenArt, responds correctly to its own transactions and handles later uninitiated transactions for the target inconsistently (ignoring, aborting or acknowledging them). Enea could not recover the SilentAir binary because of strong OPSEC. Activity arrives in bursts separated by months of inactivity, ranging from one or a few subscribers to several thousand subscribers in a month.

Observed message types: information harvesting via SS7 SRI-SM, RestoreData and SendAuthenticationInfo and Diameter Authentication-Information-Request (AIR); location tracking via SS7 AnyTimeInterrogation (ATI) and Diameter Insert-Subscriber-Data-Request (cell IDs); and interception via SS7 UpdateLocation (UL) and Diameter Update-Location-Request (ULR), used to redirect voice and SMS for targets. 2FA codes were observed immediately after interception for some targets. Activity mix per Enea: interception 18.30%, information harvesting 7.21%, location tracking 0.22%, other (mostly silent SMS) 74.27% - the inverse of a typical surveillance company profile (61.89% harvesting, 30.90% location tracking). Large volumes of silent SMS, primarily aimed at Taiwanese subscribers and submitted via interim SMSCs in Southeast Asia and Taiwan, spike around midnight and are assessed as a way to confirm interception remains in place or to bulk-test interception. Targets include political and government VIPs in Taiwan and roughly a dozen further countries and territories (Southeast/East/South/Central Asia, East Africa, the Middle East, Europe).

Enea notes a possible but unconfirmed link to LIMINAL PANDA (CrowdStrike, low-confidence China-nexus; tracked by Unit 42 as CL-STA-0969), the only actor Enea knows to actively exploit the IPX interconnect network. Evidence for a link: SilentAir activity overlapped LIMINAL PANDA infection periods in some operators, SilentAir attacks were sometimes followed by GTPDoor tunnel attempts and vice versa, and both target Asian operators; against it, LIMINAL PANDA has no known Taiwan focus and the SCTPE SS7/Diameter injector only processes self-generated transactions within a 5-second window, which cannot explain SilentAir's handling of uninitiated transactions. Enea does not associate the two groups and says they may share access or operate in parallel. Indicators and ATT&CK mappings below that derive from LIMINAL PANDA / CL-STA-0969 reporting are therefore related-actor context, not confirmed SilentAir tooling.

MITRE ATT&CK techniques by tactic:

Defense Evasion

T1014 Rootkit; T1036.005 Match Legitimate Resource Name or Location; T1070.003 Clear Command History; T1070.006 Timestomp; T1205 Traffic Signaling

Lateral Movement

T1021.004 SSH

Discovery

T1046 Network Service Discovery

Command and Control

T1071.004 DNS; T1090 Proxy; T1095 Non-Application Layer Protocol; T1572 Protocol Tunneling

Initial Access

T1078 Valid Accounts; T1199 Trusted Relationship

Credential Access

T1110.001 Password Guessing; T1111 Multi-Factor Authentication Interception; T1557 Adversary-in-the-Middle; T1557.001 Name Resolution Poisoning and SMB Relay

Persistence

T1556.003 Pluggable Authentication Modules

Resource Development

T1584 Compromise Infrastructure; T1584.008 Compromise Infrastructure

Remediation

Immediate actions

  • Deploy SS7 and Diameter signaling firewalls with cross-protocol correlation and alert on UpdateLocation/Update-Location-Request, AnyTimeInterrogation, SRI-SM, SendAuthenticationInfo and AIR originating from unexpected or internal-looking Global Titles
  • Hunt for bursts of silent SMS and unusual SMSC submissions targeting individual subscribers, especially high-profile political and government subscribers
  • Treat the IPX/GRX as untrusted: monitor GTP-C (UDP 2123) and GTP-U traffic, DNS and ICMP crossing the interconnect for covert C2
  • Hunt for GTPDoor v3 (MD5 35f0d4d0c5519a2957338e448fce53bc) and SCTPE (MD5 da35d98c9a22e8ede93595d21fe7e63e) on Linux core-network hosts

Workarounds

  • High-risk subscribers should avoid SMS-based 2FA and use app- or hardware-token-based authentication
  • Prefer end-to-end encrypted messaging and calling over SMS/voice for sensitive communications

Longer-term hardening

  • Go beyond baseline GSMA signaling-security recommendations: add GTP-C/GTP-U security aligned with SS7/Diameter firewalls
  • Build detection logic for internal signaling protocols such as SIGTRANslator and Cordscan traffic even where firewalling is not feasible
  • Integrate signaling security telemetry with broader IT security monitoring
  • Assess the effect of outsourced core networks on investigations and digital sovereignty
  • Harden Linux telecom nodes: remove default accounts, enforce key-based SSH, patch legacy privilege-escalation flaws

Affected and fixed versions

  • Various mobile network operators — SS7 / Diameter signaling nodes and core-network hosts
    Vulnerable versions: Operator nodes in at least six Asian countries compromised and used as signaling attack sources

Timeline

Last updated · revalidated 1× · latest source

  • GTPDoor first publicly analysed by HaxRob (two versions); later tied to LIMINAL PANDA / CL-STA-0969 activity.
  • CrowdStrike assesses LIMINAL PANDA telecom intrusion activity as China-nexus with low confidence.
  • Enea begins observing unusual SS7 and Diameter activity from Asian operators attributed to SilentAir (early 2025); follow-up analysis finds earlier activity.
  • Palo Alto Unit 42 publishes CL-STA-0969 report on Southwest/Southeast Asian telecom intrusions (Feb-Nov 2024), assessed as heavily overlapping LIMINAL PANDA.
  • From Q1 2026 Enea detects and blocks GTP-C GTPDoor activity at several customer operators (several hundred covert C2 attempts); 5-10 operators estimated infected, possibly several dozen globally.
  • CrowdStrike identifies SIGweaver (SCTPE), a Linux kernel-level SS7/Diameter injection driver, and links it to Liminal Panda with moderate confidence (per IMDA advisory).
  • POST Luxembourg publicly reports GTP-C GTPDoor activity detected on its network.
  • Enea publishes 'The Enemy Within' naming SilentAir as a new state-level signaling actor and discussing a possible, unconfirmed link to LIMINAL PANDA (updated 2026-11-12).
  • Enea updates 'The Enemy Within' with Key Takeaways.
  • The Hacker News reports SilentAir as disrupted for now but able to return; Enea has notified the operators where attack sources were found.
  • SecurityWeek (Eduard Kovacs) reports SilentAir intercepting calls and SMS from inside mobile networks.

Update history

Sources cited

Drafted by the Threadlinqs research pipeline from 7 sources; published ; last updated .

Threadlinqs Intelligence — Real-Time Threat Detection Platform

// threat_feed
$ sort --newest
Showing all threats

Live intelligence console

Threat level

Latest Threats