This OTX pulse (AlienVault, TLP:WHITE, modified 2026-08-26) documents a phishing operation that began as a Canada Revenue Agency (CRA) T4 tax-document lure and has since scaled into a global campaign spanning 46 countries, with roughly 45% of observed activity concentrated in the United States. Targeted verticals include Education, Technology, Government, Finance, and Manufacturing across the US and Canada.
The defining characteristic of this campaign is that it does not drop traditional malware. Instead, attackers impersonate trusted institutions — the Canada Revenue Agency, the US Social Security Administration, Adobe, and regional tax authorities — to trick victims into installing legitimate Remote Monitoring & Management (RMM) software. Once installed, the RMM agent grants operators persistent, hands-on-keyboard access that blends into normal administrative traffic. This is a textbook living-off-the-land (LotL) tradecraft: the 'payload' is signed, trusted software, so signature-based AV and reputation filters are largely blind to it.
Delivery relies on password-protected archives attached to or linked from phishing emails. The password is provided in the lure body (often styled as a 'document protection' measure), which defeats email gateway sandbox detonation — sandboxed extractors cannot open the archive without the password. Infrastructure analysis shows a blend of attacker-registered throwaway domains (e.g., quavix.vu, cevora.vu, xorlira.vu on the .vu ccTLD), abused dynamic-DNS style hosts (swoop2.me, letsgo2.me), and Vercel-hosted phishing pages that inherit the trust and TLS legitimacy of a major cloud platform.
The adversary's objective is persistent remote access for follow-on monetization: data theft, business email compromise staging, credential harvesting, and resale of access to ransomware affiliates or initial access brokers. Hands-on-keyboard RMM access is consistently one of the top precursors to enterprise ransomware detonation.
Threat Actor / Malware Profile
Attribution: Unknown threat actor. No named malware family — the 'malware' is commercially legitimate RMM tooling (e.g., ScreenConnect/ConnectWise, AnyDesk, TeamViewer-class agents) installed under attacker-controlled configurations.
Distribution method:
- Spear-phishing emails impersonating CRA (T4 tax documents), SSA, Adobe, and tax authorities
- Password-protected ZIP archives to evade sandbox extraction and gateway inspection
- Phishing pages hosted on Vercel infrastructure and look-alike domains to proxy victim trust
Payload behavior:
- Victim executes an RMM installer (often an MSI or signed EXE) extracted from the protected archive
- Installer registers the endpoint with attacker-controlled RMM tenant/instance IDs
- Operator gains interactive remote session — full hands-on-keyboard capability
- Follow-on activity typically includes browser credential theft, document staging, and internal reconnaissance
C2 communication:
- Outbound TLS to legitimate RMM vendor infrastructure (vendor-cloud relay) or attacker-hosted relay hosts such as
dcsi23.swoop2.meandssi11.letsgo2.me - Because traffic rides vendor cloud services, network detection must rely on which RMM processes connect and which relay/hostnames are involved, not on protocol anomaly alone
Persistence mechanism:
- RMM agent installs as a Windows service (e.g., services created by ScreenConnect/AnyDesk-class installers) and/or Run-key persistence, surviving reboots
- Agent auto-reconnects to the attacker tenant on startup
Anti-analysis techniques:
- Password-protected archives defeating automated detonation
- Signed, legitimate binaries evading reputation and signature engines
- Use of reputable hosting (Vercel) and TLS for lure pages
- Rapid domain churn across throwaway
.vuregistrations and dynamic-DNS subdomains
IOC Analysis
The pulse carries 35 indicators, dominated by hostname and domain types — consistent with an infrastructure-centric campaign rather than hash-detectable malware:
- Hostnames (
dashboarduat.paynnow.com,dcsi23.swoop2.me,ssi11.letsgo2.me): likely phishing portals and RMM relay/staging endpoints. These are the highest-fidelity indicators — enumerate full DNS and proxy logs for exact-match hits. - Apex domains (
swoop2.me,hiltonheadislanddeals.com,quavix.vu,cevora.vu,xorlira.vu): wildcard-block at DNS sinkhole / secure web gateway. Dynamic-DNS and compromised-legitimate domains (hiltonheadislanddeals.com) warrant subdomain-level logging rather than assuming the apex is the malicious node.
Operationalization guidance for SOC teams:
- Push all domains/hostnames into DNS RPZ, SWG blocklists, and EDR custom IOC lists with 90-day retroactive lookback — RMM access often sits dormant for weeks before activation.
- Because indicators are domain-only, pair IOC matching with behavioral coverage (see Detection Engineering): any unauthorized RMM process on the endpoint is a finding regardless of infrastructure.
- Enrich hits with passive DNS (e.g., SecurityTrails, VirusTotal, any.run sandbox reports from the pulse reference) to catch infrastructure rotation — actors behind dynamic-DNS hosts re-point subdomains frequently.
- any.run interactive sandbox remains the fastest way to detonate the password-protected archives manually using the lure-supplied password to extract dropped installer hashes and RMM tenant IDs for additional pivots.
Detection Engineering
---
title: Unauthorized RMM Tool Execution (ScreenConnect / AnyDesk / TeamViewer Class)
id: 7f3a2c1e-9b4d-4e6a-a1c2-rmmphish0001
status: experimental
description: Detects execution of commonly abused legitimate RMM binaries consistent with phishing-delivered RMM campaigns (CRA T4 / SSA lure tradecraft). Tune allowlist to your approved RMM tooling.
author: Security Arsenal Threat Intel
references:
- https://any.run/cybersecurity-blog/us-campaign-malware-analysis
date: 2026/08/26
logsource:
category: process_creation
product: windows
detection:
selection_img:
Image|endswith:
- '\screenconnect.clientservice.exe'
- '\screenconnect.windowsclient.exe'
- '\anydesk.exe'
- '\teamviewer.exe'
- '\teamviewer_service.exe'
- '\rustdesk.exe'
- '\ammyy.exe'
- '\aeroadmin.exe'
- '\splashtop.exe'
- '\netop.exe'
filter_approved_paths:
Image|startswith:
- 'C:\Program Files\ApprovedRMM\'
condition: selection_img and not filter_approved_paths
falsepositives:
- Legitimate IT helpdesk RMM usage — maintain an approved-path/process allowlist
level: high
tags:
- attack.command_and_control
- attack.t1219
---
title: Password-Protected Archive Extraction Followed by Installer Execution
id: 8a4b3d2f-1c5e-4f7b-b2d3-rmmphish0002
status: experimental
description: Detects extraction utilities or Explorer spawning MSI/installer processes shortly after archive handling — pattern typical of password-protected ZIP phishing lures delivering RMM installers.
author: Security Arsenal Threat Intel
references:
- https://any.run/cybersecurity-blog/us-campaign-malware-analysis
date: 2026/08/26
logsource:
category: process_creation
product: windows
detection:
selection_parent:
ParentImage|endswith:
- '\explorer.exe'
- '\7z.exe'
- '\winrar.exe'
- '\winzip.exe'
selection_child:
Image|endswith:
- '\msiexec.exe'
- '\setup.exe'
- '\install.exe'
selection_cmd:
CommandLine|contains:
- 'Downloads'
- 'Temp'
- 'AppData\Local\Temp'
condition: selection_parent and selection_child and selection_cmd
falsepositives:
- User-installed software from Downloads — correlate with email/proxy telemetry for phishing context
level: medium
tags:
- attack.initial_access
- attack.t1566.001
- attack.execution
---
title: Suspicious Outbound Connection to Known RMM Phishing Infrastructure
id: 9c5d4e3a-2d6f-5a8c-c3e4-rmmphish0003
status: experimental
description: Detects DNS/network connections to infrastructure identified in the OTX RMM phishing pulse, including dynamic-DNS relays and throwaway .vu domains.
author: Security Arsenal Threat Intel
date: 2026/08/26
logsource:
category: dns
product: windows
detection:
selection:
query|contains:
- 'dashboarduat.paynnow.com'
- 'dcsi23.swoop2.me'
- 'ssi11.letsgo2.me'
- 'swoop2.me'
- 'hiltonheadislanddeals.com'
- 'quavix.vu'
- 'cevora.vu'
- 'xorlira.vu'
condition: selection
falsepositives:
- None expected; these are campaign-specific indicators
level: critical
tags:
- attack.command_and_control
- attack.t1071.001
// Hunt: RMM phishing campaign — IOC matches + unauthorized RMM process/network behavior
let lookback = 90d;
let iocs = dynamic(["dashboarduat.paynnow.com","dcsi23.swoop2.me","ssi11.letsgo2.me","swoop2.me","hiltonheadislanddeals.com","quavix.vu","cevora.vu","xorlira.vu"]);
let rmmProcs = dynamic(["screenconnect.clientservice.exe","screenconnect.windowsclient.exe","anydesk.exe","teamviewer.exe","teamviewer_service.exe","rustdesk.exe","splashtop.exe","aeroadmin.exe","netop.exe"]);
let NetHits = DeviceNetworkEvents
| where Timestamp > ago(lookback)
| where RemoteUrl has_any (iocs)
| project Timestamp, DeviceName, InitiatingProcessFileName, InitiatingProcessCommandLine, RemoteUrl, RemoteIP, ActionType;
let ProcHits = DeviceProcessEvents
| where Timestamp > ago(lookback)
| where FileName in~ (rmmProcs)
| project Timestamp, DeviceName, FileName, ProcessCommandLine, FolderPath, InitiatingProcessFileName, InitiatingProcessCommandLine, SHA256;
let RecentRMMInstalls = DeviceEvents
| where Timestamp > ago(lookback)
| where ActionType == "ServiceInstalled"
| extend ServiceName = tostring(parse_json(AdditionalFields).ServiceName)
| where ServiceName has_any ("screenconnect","anydesk","teamviewer","rustdesk","splashtop")
| project Timestamp, DeviceName, ServiceName, InitiatingProcessFileName;
union NetHits, ProcHits
| sort by Timestamp desc;
RecentRMMInstalls
| sort by Timestamp desc
# Security Arsenal - RMM Phishing Campaign IOC & Artifact Hunt (Windows endpoints)
# Run via EDR Live Response / RMM-of-record / GPO. Requires admin for service & all-user checks.
$iocDomains = @('dashboarduat.paynnow.com','dcsi23.swoop2.me','ssi11.letsgo2.me','swoop2.me','hiltonheadislanddeals.com','quavix.vu','cevora.vu','xorlira.vu')
$rmmNames = @('ScreenConnect','AnyDesk','TeamViewer','RustDesk','AeroAdmin','Splashtop','NetSupport','Netop')
$findings = @()
# 1) Unauthorized RMM services (persistence)
Get-CimInstance Win32_Service | Where-Object {
$n = $_.Name + ' ' + $_.DisplayName + ' ' + $_.PathName
$rmmNames | Where-Object { $n -match $_ }
} | ForEach-Object {
$findings += [pscustomobject]@{Type='Service'; Detail="$($_.Name) | $($_.DisplayName) | $($_.PathName) | State=$($_.State)"}
}
# 2) Run-key persistence for RMM agents (HKLM + all user hives)
$runKeys = @('HKLM:\SOFTWARE\Microsoft\Windows\CurrentVersion\Run',
'HKLM:\SOFTWARE\Microsoft\Windows\CurrentVersion\RunOnce',
'HKCU:\SOFTWARE\Microsoft\Windows\CurrentVersion\Run',
'HKCU:\SOFTWARE\Microsoft\Windows\CurrentVersion\RunOnce')
foreach ($rk in $runKeys) {
if (Test-Path $rk) {
(Get-ItemProperty $rk).PSObject.Properties | Where-Object {
$rmmNames | Where-Object { "$($_.Value)" -match $_ }
} | ForEach-Object {
$findings += [pscustomobject]@{Type='RunKey'; Detail="$rk :: $($_.Name) = $($_.Value)"}
}
}
}
# 3) Running RMM processes
Get-Process | Where-Object { $rmmNames | Where-Object { $_.Name -match "^$_" } } | ForEach-Object {
$findings += [pscustomobject]@{Type='Process'; Detail="$($_.Name) (PID $($_.Id)) Path=$($_.Path)"}
}
# 4) DNS cache & active connections matching campaign IOCs
$dnsCache = Get-DnsClientCache -ErrorAction SilentlyContinue
foreach ($d in $iocDomains) {
$hits = $dnsCache | Where-Object { $_.Entry -like "*$d*" }
foreach ($h in $hits) { $findings += [pscustomobject]@{Type='DNSCache'; Detail="$($h.Entry) -> $($h.Data)"} }
}
Get-NetTCPConnection -State Established -ErrorAction SilentlyContinue | ForEach-Object {
try {
$proc = Get-Process -Id $_.OwningProcess -ErrorAction Stop
if ($rmmNames | Where-Object { $proc.Name -match $_ }) {
$findings += [pscustomobject]@{Type='NetConnection'; Detail="$($proc.Name) (PID $($_.OwningProcess)) -> $($_.RemoteAddress):$($_.RemotePort)"}
}
} catch {}
}
# 5) Recent MSI installs of RMM products (install artifact)
Get-CimInstance Win32_Product -ErrorAction SilentlyContinue | Where-Object {
$rmmNames | Where-Object { $_.Name -match $_ }
} | ForEach-Object {
$findings += [pscustomobject]@{Type='InstalledProduct'; Detail="$($_.Name) | InstallDate=$($_.InstallDate)"}
}
if ($findings.Count -gt 0) {
Write-Host "[ALERT] $($findings.Count) potential RMM-phishing artifacts on $env:COMPUTERNAME" -ForegroundColor Red
$findings | Format-Table -AutoSize
$findings | Export-Csv -NoTypeInformation -Path "$env:TEMP\rmm_hunt_$env:COMPUTERNAME.csv"
} else {
Write-Host "[CLEAN] No RMM-phishing artifacts detected on $env:COMPUTERNAME" -ForegroundColor Green
}
Response Priorities
Immediate (0–4h):
- Block all pulse IOCs (hostnames + apex domains) at DNS sinkhole, secure web gateway, and EDR custom indicators; wildcard-block
*.swoop2.me,*.letsgo2.me, and the.vuthrowaway domains pending business justification. - Run the KQL hunt and PowerShell sweep fleet-wide with a 90-day lookback; any unauthorized RMM service or process is treated as an active intrusion, not a policy violation — isolate the host and capture memory before remediation.
- Block inbound email carrying password-protected archives where the password appears in the message body (a near-universal phishing tell) and flag Vercel-hosted links to tax/SSA-themed lures.
24 hours:
- For any host with confirmed RMM agent presence, assume full credential compromise: force password resets for all users who authenticated on that host, revoke active sessions/tokens, and rotate any local admin or service account credentials the operator could have observed on-screen.
- Review VPN, M365, and SSO logs for anomalous logins sourced from the compromised window of access; hands-on-keyboard operators frequently harvest browser credential stores and session cookies within the first hours.
- Inspect outbound data volumes from affected endpoints for staging/exfiltration and check for newly created local accounts or modified RDP/firewall settings.
1 week:
- Implement an RMM allowlist architecture: application control (WDAC/AppLocker) permitting only the organization's sanctioned RMM product, with alerting on any other remote-access binary or service installation.
- Deploy the Sigma rules above to the SIEM and enable service-installation auditing (Event ID 4697/7045) forwarding from all endpoints.
- Tune email security to detonate password-protected archives with supplied passwords extracted from message bodies, and launch targeted user awareness on tax/SSA/Adobe-themed lures for Finance, HR, and Education-sector staff.
- Add retroactive threat-hunt standing queries for dynamic-DNS and
.vu-TLD RMM relay patterns to catch infrastructure rotation as the actor re-registers domains.
Related Resources
Security Arsenal Incident Response Managed SOC & MDR Services AlertMonitor Threat Detection From The Dark Side Intel Hub
Is your security operations ready?
Get a free SOC assessment or see how AlertMonitor cuts through alert noise with automated triage.