Skip to navigation

Threat reportMalwareTL-2026-1512

UNC5142 EtherHiding: BNB Smart Chain-Based Malware Distribution via Compromised WordPress Sites

highTRACKING

UNC5142, a financially motivated cluster tracked by Mandiant Threat Defense and Google Threat Intelligence Group since late 2023, abused a three-level BNB Smart Chain smart-contract system ("EtherHiding") to dynamically store and serve malicious CLEARSHORT JavaScript payload configurations across roughly 14,000 compromised WordPress sites, using evolving social-engineering ClickFix-style lures to deliver VIDAR, LUMMAC.V2 (Lumma), RADTHIEF (Rhadamanthys), and ATOMIC (AMOS) infostealers to Windows and macOS victims.

Key facts

IdentifiersTL-2026-1512
Also known asEtherHiding Campaign, CLEARSHORT
SeverityHIGH
Affected productsWordPress WordPress core / plugins / themes (vulnerable/outdated installations); Multiple Windows end-user endpoints (infostealer targets: VIDAR, LUMMAC.V2, RADTHIEF); Apple macOS end-user endpoints (infostealer target: ATOMIC/AMOS)
AttributionUNC5142 (high confidence)
MalwareATOMIC, CLEARSHORT, LUMMAC.V2, RADTHIEF, Vidar
Tools0x53fd54f55C93f9BCCA471cD0CcbaBC3Acbd3E4AA, 0x8FBA1667BEF5EdA433928b220886A830488549BD, 0x9179dda8B285040Bf381AABb8a1f4a1b8c37Ed53
MITRE ATT&CK techniquesT1005 Data from Local System, T1027 Obfuscated Files or Information, T1036 Masquerading, T1041 Exfiltration Over C2 Channel, T1059.001 PowerShell, T1070 Indicator Removal, T1071.001 Web Protocols, T1082 System Information Discovery, T1140 Deobfuscate/Decode Files or Information, T1190 Exploit Public-Facing Application, T1202 Indirect Command Execution, T1204.002 Malicious File and 11 more in the attack flow
MotivationFINANCIAL
Target sectorstechnology, ecommerce, general internet users, cryptocurrency
Target regionsGlobal
Detection rules9 (1 shown free)
Sources9 cited sources

Detection coverage

As of 2026-01-01, Threadlinqs Intelligence publishes 9 detection rule(s) for TL-2026-1512 across Splunk SPL, Microsoft KQL and Sigma, covering 42 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: ClickFix mshta.exe Spawning PowerShell with CLEARSHORT-Style Flags (Post-Exploitation)

Detects the Windows post-exploitation execution chain observed in UNC5142's ClickFix lure flow: mshta.exe launched from a user-pasted Run-dialog command spawning powershell.exe with the specific flag pattern (-ep RemoteSigned -w 1 -enc) used to decrypt and execute the CLEARSHORT PowerShell loader in memory.

  • Languages: Splunk SPL, Microsoft KQL, Sigma
  • ATT&CK: T1218.005, T1059.001, T1204.002
  • Data sources: sysmon, windows_security, edr_process_events
  • Severity: critical
  • Confidence: high

Splunk SPL

index=sysmon sourcetype="XmlWinEventLog:Microsoft-Windows-Sysmon/Operational" EventCode=1
  ParentImage="*\\mshta.exe" Image="*\\powershell.exe"
  CommandLine="*-ep RemoteSigned*" CommandLine="*-w 1*" CommandLine="*-enc*"
| stats count min(_time) as first_seen values(CommandLine) as command_lines by ComputerName, User, ParentImage, Image
| where count > 0
| table first_seen, ComputerName, User, ParentImage, Image, command_lines

Microsoft KQL

DeviceProcessEvents
| where Timestamp > ago(24h)
| where InitiatingProcessFileName =~ "mshta.exe" and FileName =~ "powershell.exe"
| where ProcessCommandLine has "-ep RemoteSigned" and ProcessCommandLine has "-w 1" and ProcessCommandLine has "-enc"
| project Timestamp, DeviceName, AccountName, InitiatingProcessFileName, InitiatingProcessCommandLine, FileName, ProcessCommandLine
| summarize count(), make_set(ProcessCommandLine) by DeviceName, AccountName

Sigma

title: Suspicious mshta.exe Spawning PowerShell with CLEARSHORT-style Flags
id: 8b2f1a5e-1512-4a1f-9c1e-clearshort01
status: experimental
description: Detects mshta.exe spawning powershell.exe with the -ep RemoteSigned -w 1 -enc flag pattern used by UNC5142's CLEARSHORT ClickFix-delivered PowerShell loader to decrypt and run second-stage code in memory.
references:
    - https://cloud.google.com/blog/topics/threat-intelligence/unc5142-etherhiding-distribute-malware/
author: Threadlinqs Team
date: 2026-07-19
tags:
    - attack.execution
    - attack.t1218.005
    - attack.t1059.001
    - attack.t1204.002
    - attack.tool.clearshort
logsource:
    category: process_creation
    product: windows
detection:
    selection_parent:
        ParentImage|endswith: '\mshta.exe'
    selection_child:
        Image|endswith: '\powershell.exe'
        CommandLine|contains|all:
            - '-ep RemoteSigned'
            - '-w 1'
            - '-enc'
    condition: selection_parent and selection_child
falsepositives:
    - Legitimate administrative scripts or HTA-based internal tooling that legitimately invoke PowerShell with similar flags (rare in modern environments)
level: high

Known false positives:

  • Legacy internal HTA-based admin tooling that legitimately chains mshta.exe to powershell.exe with a similar flag set (rare — verify against a change-managed allowlist)

Tuning notes: Treat any hit as high-confidence unless the parent mshta.exe process maps to a known, change-managed internal HTA tool; correlate with a preceding browser process (chrome.exe/msedge.exe) launching mshta.exe via Run dialog for near-certain ClickFix confirmation.

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)
  • WordPress Theme/Plugin File Modification Consistent with CLEARSHORT Content Injection (Exploitation) · unlock with TL Creds
  • Mark-of-the-Web / Quarantine Bypass on Recently-Downloaded .hta/.xll File (Post-Exploitation) · unlock with TL Creds
  • Legitimate-Service Payload Staging Abuse (Cloudflare Pages / MediaFire / GitHub Raw) Preceding Infostealer Execution (C2 / Lateral Movement) · unlock with TL Creds
  • PowerShell In-Memory .NET Loader Decryption Chain Post-CLEARSHORT (Exploitation → Execution Bridge) · unlock with TL Creds
  • Suspicious HKCU Run Key Persistence Correlated with ClickFix-Style Browser Prompt (Post-Exploitation) · unlock with TL Creds
  • macOS Quarantine (com.apple.quarantine) Removal on Recently-Downloaded File via xattr (C2 / Defense-Evasion Bridge) · unlock with TL Creds
  • EtherHiding CLEARSHORT Web3.js Load + eth_call to bsc-dataseed (Exploitation) · unlock with TL Creds
  • STUN-Based IP Fingerprinting Followed by POST to Newly-Registered Domain (C2 / Recon Beacon) · unlock with TL Creds

Indicators of compromise

As of 2026-01-01, Threadlinqs Intelligence tracks 42 indicators of compromise for TL-2026-1512.

Sample indicators (3 of 42)

  • IP address: 91.240.118[.]2
  • Domain: saaadnesss[.]shop
  • File hash: bcbdb74f97092dfd68e7ec1d6770b6d1e1aae091f43bcebb0b7bce6c8188e310

Defanged, TLP:CLEAR. Full set of 42 indicators: unlock with TL Creds.

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

Technical analysis

No activity has been observed since July 23, 2025.

UNC5142 compromises WordPress sites running vulnerable core, plugin, or theme versions and injects a multistage JavaScript downloader family tracked as CLEARSHORT into theme files (header.php, footer.php, index.php), plugin directories, or the WordPress database. On page load, CLEARSHORT Stage 1 loads the Web3.js, pako (gzip), and crypto-js libraries and connects to the BNB Smart Chain via the public bsc-dataseed.binance.org RPC node to query a First-Level (router) smart contract, which returns a Base64/gzip-encoded ABI and the address of a Second-Level (logic) smart contract. The Second-Level contract exposes four functions used for victim fingerprinting and reconnaissance -- teaCeremony (dynamic code execution / POST check-ins), shibuyaCrossing (OS/platform identification), asakusaTemple (interaction beaconing), and ginzaLuxury (malicious lure retrieval/decryption) -- and its returned script is directly eval()'d in the victim browser. That script performs WebRTC/STUN-based IP address recovery (querying stun.l.google.com:19302 and POSTing the recovered IP to actor-controlled check-in domains such as saaadnesss[.]shop and lapkimeow[.]icu, later ratatui[.]today) and queries a Third-Level (storage) smart contract that holds an AES-GCM-encrypted CLEARSHORT landing page URL, the AES key, and second-stage payload URLs. The decrypted landing page presents an evolving series of ClickFix-style social-engineering lures -- fake Chrome update prompts, fake reCAPTCHA/Data Privacy dialogs, a spoofed Cloudflare "Unusual Web Traffic" error, and later an "Anti-Bot Verification" prompt for both Windows and macOS -- that trick the victim into executing an attacker-supplied command. This leads to a four-stage delivery chain: an initial dropper (.hta or a .xll file masquerading as an Excel add-in) fetched from Cloudflare Pages (*.pages.dev) or attacker infrastructure; a PowerShell loader (invoked with -ep RemoteSigned -w 1 -enc) that performs AES/TripleDES decryption and defense evasion (including ipconfig /flushdns and Mark-of-the-Web bypasses via xattr -c on macOS or NTFS Zone.Identifier stream removal on Windows); abuse of legitimate services (GitHub, MediaFire, Cloudflare Pages, and in early campaigns Backblaze B2) to host an encrypted payload blob disguised with benign extensions (.mp4, .mp3, .wav, .dat); and finally in-memory, no-disk-write execution of a .NET loader that decrypts and runs the final infostealer payload. UNC5142 operated two parallel smart-contract infrastructures -- a Main system deployed November 24, 2024 and a Secondary system deployed February 18, 2025 -- both funded from the same OKX exchange intermediary wallet and updated in near-lockstep (including a coordinated update on March 3, 2025), which GTIG assesses with high confidence indicates single-actor control. Updates to the on-chain configuration typically cost $0.25-$1.50 in BNB network fees, making the infrastructure extremely cheap to maintain and, because the blockchain component is immutable and publicly distributed, effectively resistant to conventional takedown. GTIG does not attribute the final infostealer payloads to UNC5142 directly and assesses the cluster most likely operates as a stolen-credential/data distribution service selling access to downstream threat actors. No CVE is associated with this campaign; initial access is achieved purely through vulnerable WordPress installations and social engineering rather than a specific software vulnerability. No UNC5142 activity has been observed since July 23, 2025, which GTIG assesses may represent an operational pause or a shift in tradecraft rather than the end of the cluster.

MITRE ATT&CK techniques by tactic:

Collection

T1005 Data from Local System

Defense Evasion

T1027 Obfuscated Files or Information; T1036 Masquerading; T1070 Indicator Removal; T1140 Deobfuscate/Decode Files or Information; T1202 Indirect Command Execution

Exfiltration

T1041 Exfiltration Over C2 Channel

Execution

T1059.001 PowerShell; T1204.002 Malicious File

Command and Control

T1071.001 Web Protocols; T1219 Remote Access Tools; T1568 Dynamic Resolution

Discovery

T1082 System Information Discovery; T1614 System Location Discovery

Initial Access

T1190 Exploit Public-Facing Application

stealth

T1216 System Script Proxy Execution; T1218.005 Mshta

Credential Access

T1539 Steal Web Session Cookie; T1552.001 Credentials In Files

defense-impairment

T1553.005 Mark-of-the-Web Bypass

Persistence

T1554 Compromise Host Software Binary

discovery

T1580 Cloud Infrastructure Discovery

Resource Development

T1583.006 Web Services

Remediation

Patches

  • Apply the latest official security patches/updates for the specific WordPress core version, theme, and plugins in use (no single CVE identified for this campaign)

Immediate actions

  • Block all known UNC5142 C2 and payload-hosting domains and IPs at DNS/network egress
  • Update WordPress core, themes, and plugins on all internet-facing sites immediately
  • Deploy file integrity monitoring on WordPress theme/plugin directories and the WordPress database
  • Alert on mshta.exe execution and on powershell.exe spawned from mshta.exe with -ep RemoteSigned -w 1 -enc in the command line
  • Alert on outbound connections to mediafire.com/file_premium and suspicious GitHub raw-content downloads

Workarounds

  • Where immediate patching is not possible, place vulnerable WordPress plugin/theme endpoints behind WAF virtual patching rules
  • Disable or remove unused/outdated WordPress plugins and themes that are common compromise vectors

Longer-term hardening

  • Implement a Web Application Firewall (WAF) tuned to detect and block malicious JavaScript injection on WordPress
  • Restrict PHP execution in wp-content/plugins and wp-content/themes directories
  • Deploy WordPress security/hardening plugins (e.g. Wordfence, Sucuri) with active malware scanning
  • Monitor for Web3.js library loads (cdn.jsdelivr.net/npm/web3) and eth_call RPC traffic to bsc-dataseed.binance.org as a web-layer detection signal for EtherHiding
  • Monitor RTCPeerConnection/STUN usage (stun.l.google.com:19302) on customer-facing web properties as a fingerprinting-evasion indicator
  • Monitor and restrict abuse of Cloudflare Pages (*.pages.dev), MediaFire, GitHub, and Backblaze B2 as payload-hosting channels

Affected and fixed versions

  • WordPress — WordPress core / plugins / themes (vulnerable/outdated installations)
    Vulnerable versions: Outdated or misconfigured WordPress core, plugin, and theme installations (no specific version enumerated by source)
    Fixed in: Latest patched WordPress core, plugin, and theme versions
  • Multiple — Windows end-user endpoints (infostealer targets: VIDAR, LUMMAC.V2, RADTHIEF)
    Vulnerable versions: Windows systems exposed to ClickFix-style social engineering
  • Apple — macOS end-user endpoints (infostealer target: ATOMIC/AMOS)
    Vulnerable versions: macOS systems exposed to ClickFix-style social engineering

Timeline

  • Mandiant Threat Defense and Google Threat Intelligence Group begin tracking UNC5142 as a distinct financially motivated cluster.
  • Earliest observed CLEARSHORT campaigns use a single smart contract, Base64 encoding, .shop-TLD hosting, and fake Chrome update lures.
  • Campaign adds STUN-based IP reconnaissance and fake reCAPTCHA ClickFix lures; hosting expands to Cloudflare *.pages.dev and .icu domains.
  • Main three-level smart contract infrastructure deployed on BNB Smart Chain, funded with 0.1 BNB (~$66) from an OKX exchange intermediary wallet.
  • Three-contract system refined; landing pages begin using AES-GCM encryption alongside Base64, with Data Privacy agreement and reCAPTCHA lure variants.
  • ATOMIC (AMOS) macOS infostealer first observed being distributed via the campaign; spoofed Cloudflare 'Unusual Web Traffic' error lure introduced.
  • Secondary smart contract infrastructure deployed, funded with 0.235 BNB (~$152) from the same OKX intermediary wallet that funded the Main infrastructure.
  • LUMMAC.V2 (Lumma) and RADTHIEF (Rhadamanthys) infostealers first observed being distributed alongside VIDAR.
  • Main and Secondary infrastructures receive coordinated, near-simultaneous payload and lure updates, supporting single-actor attribution.
  • Lures refined to an 'Anti-Bot Verification' prompt targeting both Windows and macOS victims.
  • Google identifies approximately 14,000 compromised WordPress web pages exhibiting UNC5142/CLEARSHORT injected-JavaScript behavior.
  • Last observed UNC5142 campaign activity; GTIG assesses this may reflect an operational pause or tradecraft shift rather than cluster shutdown.
  • Google Cloud/Mandiant/GTIG publish 'New Group on the Block: UNC5142 Leverages EtherHiding to Distribute Malware,' the primary public disclosure of the campaign, including YARA rules and IOCs.

Sources cited

More from UNC5142

Previous and next in malware

Drafted by the Threadlinqs research pipeline from 9 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