Skip to navigation

Threat reportPhishingTL-2026-2452

Platform-Aware Phishing Kits Fingerprint Devices to Deliver OS-Specific RATs and Credential Harvesters

highACTIVE

Cofense Intelligence documents phishing kits that fingerprint a victim's browser/OS/language/timezone/screen/geolocation from a single landing page and branch the payload: Windows victims receive an installer (e.g. GoogleMeet_Ti.msi) that deploys the legitimate remote-management tools ConnectWise RAT (ScreenConnect) or Itarian RAT (Comodo), sometimes via a Ninite Loader staging step, while macOS and Android victims are redirected to credential-harvesting pages spoofing Google, DocuSign, Microsoft Teams, Adobe, and Zoom. First disclosed .

Key facts

IdentifiersTL-2026-2452
SeverityHIGH
First disclosed
Affected productsMicrosoft Windows (desktop endpoints); Apple macOS; Google Android
MalwareAve Maria, Ave Maria Stealer (WarzoneRAT), Ninite Loader
ToolsComodo/Itarian-domain-blended RMM channel, ConnectWise RAT (ScreenConnect), Itarian RAT (Comodo Remote Access), ScreenConnect
MITRE ATT&CK techniquesT1027 Obfuscated Files or Information, T1036 Masquerading, T1056 Input Capture, T1059 Command and Scripting Interpreter, T1071 Application Layer Protocol, T1125 Video Capture, T1204 User Execution, T1218 System Binary Proxy Execution, T1219 Remote Access Tools, T1555 Credentials from Password Stores, T1566 Phishing, T1592 Gather Victim Host Information
MotivationFINANCIAL
Target sectorsveterinary services
Detection rules9 (1 shown free)
Sources7 cited sources

Detection coverage

As of 2026-09-11, Threadlinqs Intelligence publishes 9 detection rule(s) for TL-2026-2452 across Splunk SPL, Microsoft KQL and Sigma, covering 13 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: Brand-Impersonating MSI Installer Launched via msiexec from Browser/Mail Client

Detects msiexec.exe installing an MSI package whose filename impersonates a trusted collaboration brand (Cofense's documented example: GoogleMeet_Ti.msi) where the parent process is a web browser or mail client. This is the Windows-branch resolution of the platform-aware phishing kit: a client-side fingerprinting landing page (browser UA/OS/language/timezone/screen/geolocation) routes Windows victims to a brand-impersonating installer that deploys ConnectWise RAT or Itarian RAT.

  • Languages: Splunk SPL, Microsoft KQL, Sigma
  • ATT&CK: T1592, T1566, T1204, T1218, T1036
  • Data sources: sysmon, windows_security, edr_process_telemetry
  • Severity: high
  • Confidence: high

Splunk SPL

index=sysmon OR index=endpoint sourcetype="XmlWinEventLog:Microsoft-Windows-Sysmon/Operational" EventCode=1
| eval parent_proc=lower(ParentImage), proc=lower(Image), cl=lower(CommandLine)
| where (proc="c:\\windows\\system32\\msiexec.exe" OR proc="c:\\windows\\syswow64\\msiexec.exe")
| where match(parent_proc, "(chrome|firefox|msedge|iexplore|outlook|thunderbird)\.exe$")
| where match(cl, "(googlemeet|google_meet|gmeet|docusign|msteams|zoomapp|adobereader)[a-z0-9_\-]*\.msi")
| stats count min(_time) as first_seen max(_time) as last_seen values(cl) as command_lines by host, user, parent_proc
| where count > 0
| sort -last_seen

Microsoft KQL

DeviceProcessEvents
| where Timestamp > ago(24h)
| where FileName =~ "msiexec.exe"
| where InitiatingProcessFileName has_any ("chrome.exe","msedge.exe","firefox.exe","iexplore.exe","outlook.exe","thunderbird.exe")
| where ProcessCommandLine matches regex @"(?i)(googlemeet|google_meet|gmeet|docusign|msteams|zoomapp|adobereader)[a-z0-9_\-]*\.msi"
| project Timestamp, DeviceName, AccountName, InitiatingProcessFileName, ProcessCommandLine, FolderPath, SHA256
| order by Timestamp desc

Sigma

title: Brand-Impersonating MSI Installer Launched via msiexec from Browser or Mail Client
id: 7f2b6e1a-4c3d-4a2f-9e51-2a6b8f0c1d01
status: experimental
description: Detects msiexec.exe installing an MSI package whose filename impersonates a trusted collaboration brand (e.g. GoogleMeet_Ti.msi) launched by a browser or mail client, consistent with platform-aware phishing kits that fingerprint victims before serving a Windows RAT installer.
references:
    - https://cofense.com/blog/the-platform-you-trust-is-the-platform-they-target
author: Threadlinqs Team
date: 2026/09/11
tags:
    - attack.reconnaissance
    - attack.t1592
    - attack.initial-access
    - attack.t1566.001
    - attack.execution
    - attack.t1204.002
    - attack.t1218.007
    - attack.defense-evasion
    - attack.t1036.005
logsource:
    category: process_creation
    product: windows
detection:
    selection_proc:
        Image|endswith: '\msiexec.exe'
    selection_parent:
        ParentImage|endswith:
            - '\chrome.exe'
            - '\msedge.exe'
            - '\firefox.exe'
            - '\iexplore.exe'
            - '\outlook.exe'
            - '\thunderbird.exe'
    selection_brand:
        CommandLine|contains:
            - 'googlemeet'
            - 'google_meet'
            - 'gmeet'
            - 'docusign'
            - 'msteams'
            - 'zoomapp'
            - 'adobereader'
    filter_msi:
        CommandLine|endswith: '.msi'
    condition: selection_proc and selection_parent and selection_brand and filter_msi
falsepositives:
    - Legitimate branded MSI packages pushed through approved software deployment tooling (rare when launched directly from a browser download to msiexec)
level: high

Known false positives:

  • Legitimate brand-named MSI packages distributed via approved software deployment/patch-management tooling rather than a direct browser download

Tuning notes: Extend the brand-keyword list as new lure themes are observed (this campaign family rotates Google/Zoom/DocuSign/Teams/Adobe branding); add a hash/publisher-mismatch check (Authenticode subject not matching the claimed vendor) to cut false positives from legitimately branded installers.

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)
  • Brand-Impersonating MSI Installer Launched via msiexec from Browser/Mail Client · unlock with TL Creds
  • Unauthorized ConnectWise ScreenConnect / Itarian (Comodo) RMM Install via Staged Loader · unlock with TL Creds
  • RMM C2 Beacon to Non-Allowlisted ScreenConnect/Itarian Relay + Cross-Platform Credential/Webcam Collection · unlock with TL Creds
  • Brand-Impersonating MSI Installer Launched via msiexec from Browser/Mail Client · unlock with TL Creds
  • Unauthorized ConnectWise ScreenConnect / Itarian (Comodo) RMM Install via Staged Loader · unlock with TL Creds
  • RMM C2 Beacon to Non-Allowlisted ScreenConnect/Itarian Relay + Cross-Platform Credential/Webcam Collection · unlock with TL Creds
  • Unauthorized ConnectWise ScreenConnect / Itarian (Comodo) RMM Install via Staged Loader · unlock with TL Creds
  • RMM C2 Beacon to Non-Allowlisted ScreenConnect/Itarian Relay + Cross-Platform Credential/Webcam Collection · unlock with TL Creds

Indicators of compromise

As of 2026-09-11, Threadlinqs Intelligence tracks 13 indicators of compromise for TL-2026-2452.

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

Technical analysis

Cofense Intelligence's 2026-07-01 report "The Platform You Trust Is the Platform They Target" (authored by Max Gannon) describes an evolution from one-size-fits-all phishing toward platform-aware kits that adapt payload delivery to the victim's device. A single spearphishing link (or, in a simpler baseline variant, a malicious script archive attachment) leads to a landing page that runs a client-side fingerprinting script collecting browser User-Agent, operating system, browser type, language, local time and timezone, screen/window dimensions, and geolocation. Some operators additionally use Cloudflare User-Agent edge rules to redirect traffic by perceived OS before the victim ever reaches the malicious page.

Based on the fingerprint, the kit branches: Windows victims are served an MSI-based installer (Cofense's example: GoogleMeet_Ti.msi, masquerading as a Google Meet client) that deploys a legitimate remote-monitoring-and-management (RMM) tool repurposed as a remote access trojan -- ConnectWise RAT (ScreenConnect) or Itarian RAT (Comodo). In one documented case (a veterinary-sector target), the chain runs two fingerprinting stages before resolving to either credential phishing or a loader Cofense calls Ninite Loader, which then stages ConnectWise RAT onto the Windows host. macOS and Android victims -- platforms where the same RAT installers do not apply -- are instead redirected to simplified credential-harvesting pages impersonating Google, DocuSign, Microsoft Teams, Adobe, or Zoom login flows to capture email/password pairs directly.

The report contrasts this with a simpler, traditional infection chain (Cofense ATR 389255) that used inconsistent narratives and a malicious script archive attachment to deliver Ave Maria Stealer (aka WarzoneRAT), a commodity credential-stealing RAT with keylogging, browser-credential-store harvesting, and webcam-capture capabilities. Ave Maria/WarzoneRAT was previously disrupted by an FBI/DOJ law-enforcement action in February 2024 (IC3 CSA 240215) but continues to circulate as malware-as-a-service.

Abuse of legitimate RMM software as a post-phish remote access channel is a broader, independently corroborated trend: Cofense's own May 2025 report "New Weapon of Choice" and Red Canary's June 2026 "dual-use dilemma" post both document threat actors increasingly weaponizing signed, trusted tools such as ConnectWise ScreenConnect, NetSupport Manager, and SimpleHelp because their legitimacy and vendor-domain traffic evade both signature-based detection and platform-specific security assumptions. No CVE, exploited vulnerability, or attributed threat actor/group is identified in any of the source reporting; this is a social-engineering and living-off-the-land RMM-abuse operation, not a software vulnerability.

Cofense's stated defensive takeaway is that organizations can no longer assume a given OS or device type is inherently safer, and that campaign-level, AI-clustered detection (the article cites Cofense Vision 3.2) is needed to connect the Windows-RAT and macOS/Android-credential-phishing variants as one operation rather than triaging them as unrelated, platform-siloed emails.

MITRE ATT&CK techniques by tactic:

Defense Evasion

T1027 Obfuscated Files or Information; T1036 Masquerading

Credential Access

T1056 Input Capture; T1555 Credentials from Password Stores

Collection

T1056 Input Capture; T1125 Video Capture

Execution

T1059 Command and Scripting Interpreter; T1204 User Execution

Command and Control

T1071 Application Layer Protocol; T1219 Remote Access Tools

stealth

T1218 System Binary Proxy Execution

Initial Access

T1566 Phishing

Reconnaissance

T1592 Gather Victim Host Information

Remediation

Patches

  • Not applicable -- no CVE or software vulnerability is involved; remediation is behavioral and RMM-governance control, not patching

Immediate actions

  • Alert on or block unauthorized installation of ConnectWise ScreenConnect and Itarian (Comodo) RMM agents outside approved IT change windows
  • Detect msiexec-launched installers whose embedded product/publisher metadata is spoofed or mismatched (e.g. Google/Meet-branded MSIs such as GoogleMeet_Ti.msi) on endpoints
  • Enforce phishing-resistant MFA on Google, DocuSign, Microsoft Teams, Adobe, and Zoom accounts to blunt the macOS/Android credential-harvesting branch

Workarounds

  • Restrict outbound network access to ConnectWise ScreenConnect and Itarian/Comodo RMM cloud-relay domains to only pre-approved corporate tenant instances

Longer-term hardening

  • Adopt campaign-level phishing detection (AI clustering across related lure variants) rather than per-email or per-platform triage, since this operation deliberately branches by OS to evade platform-siloed defenses
  • Maintain an allowlist of approved RMM/remote-access tool vendors and tenant instances; alert on any RMM installer or session originating outside that allowlist
  • Deploy EDR behavioral detections for keylogging, browser-credential-store access, and webcam/video capture consistent with Ave Maria (WarzoneRAT)-class stealers

Affected and fixed versions

  • Microsoft — Windows (desktop endpoints)
    Vulnerable versions: All versions -- targeted via social engineering/RMM abuse, not a software vulnerability
  • Apple — macOS
    Vulnerable versions: All versions -- redirected to credential-harvesting pages, not a software vulnerability
  • Google — Android
    Vulnerable versions: All versions -- redirected to credential-harvesting pages, not a software vulnerability

Timeline

  • MITRE ATT&CK catalogs WarzoneRAT (Ave Maria) as Software S0670, documenting the keylogging, browser-credential-theft, and webcam-capture capabilities later referenced for the stealer payload in this campaign family's traditional variant.
  • Splunk publishes 'Defending the Gates,' detection research on Ave Maria (Warzone) RAT behaviors relevant to the credential-harvesting payload used in the traditional (pre-platform-aware) infection chain.
  • FBI/DOJ disrupt Warzone RAT (Ave Maria) infrastructure per IC3 Cyber Security Advisory 240215, temporarily degrading the malware-as-a-service offering that continues to circulate in phishing kits.
  • Cofense publishes 'New Weapon of Choice,' documenting the broader trend of threat actors hijacking legitimate RMM tools including ConnectWise ScreenConnect as post-phish remote access -- the same abuse pattern used on this campaign's Windows branch.
  • Red Canary publishes 'The dual-use dilemma,' independently corroborating a surge in RMM abuse (ScreenConnect, NetSupport Manager, SimpleHelp) as ransomware/RAT delivery vectors that evade signature-based and platform-specific detection.
  • Security Boulevard syndicates the Cofense report, widening distribution of the platform-aware phishing findings to the broader security community.
  • Cofense Intelligence publishes 'The Platform You Trust Is the Platform They Target,' detailing platform-aware phishing kits (ATRs 389255, 411478, 413111/41211, 413611, 411787) that fingerprint victim devices and branch delivery to ConnectWise RAT/Itarian RAT (via Ninite Loader) on Windows or credential-harvesting pages on macOS/Android.

Sources cited

Previous and next in phishing

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