General Electric and Philips have both confirmed they are investigating claims by the Clop ransomware/extortion operation that the gang breached their systems and stole corporate data. This follows Clop's well-established playbook: compromise a widely deployed enterprise platform or file-transfer pathway, exfiltrate large volumes of sensitive data quietly, then publicly name victims on their leak site to pressure payment — often without ever deploying encryption.
For defenders, this story matters on two levels. First, if your organization operates in the GE or Philips supply chains — healthcare providers running Philips medical devices, energy and aviation firms tied to GE — you need to assess third-party exposure immediately, because Clop victim lists historically expand beyond the initial named targets. Second, Clop's campaigns are repeatable and pattern-based. The behaviors they rely on (mass data staging, archive creation, bulk outbound transfer, exploitation of internet-facing file transfer infrastructure) are all detectable if you know where to look.
This post breaks down what we know, how Clop-style intrusions typically unfold, and gives you concrete Sigma, KQL, and Velociraptor detections plus a verification script you can run today.
Technical Analysis
What Happened
- The Clop gang added General Electric and Philips to its data leak site, claiming to have stolen data from both organizations.
- Both companies publicly confirmed they are investigating the claims — notably, neither has confirmed the breach itself. This is consistent with the early phase of Clop extortion events, where the gang names victims before (or regardless of whether) negotiations occur.
- Clop's historical modus operandi centers on exploiting vulnerabilities in managed file transfer (MFT) platforms and internet-facing enterprise applications to gain initial access, then pivoting to bulk data theft. Their campaigns have repeatedly targeted file transfer appliances and enterprise software at scale, hitting hundreds of downstream victims per campaign.
Attack Chain (Defender's View)
Clop intrusions typically follow a consistent, observable chain:
- Initial Access (T1190 — Exploit Public-Facing Application): Exploitation of an internet-facing MFT platform, web application, or appliance. In past campaigns, Clop has moved from vulnerability disclosure to mass exploitation within days — speed of patching internet-facing file transfer infrastructure is the single most important control.
- Persistence / Webshell Deployment (T1505.003): Dropped webshells or backdoored components on the compromised appliance to maintain access across patching cycles.
- Discovery & Collection (T1083, T1560.001): Enumeration of accessible file shares and databases, followed by staging of stolen data into compressed archives — frequently using legitimate tools like
rar.exe,7z.exe, orwinzipwith password protection. - Exfiltration (T1567.002 / T1048): Bulk transfer of archives over HTTPS to attacker-controlled cloud storage or dedicated exfil infrastructure. Volumes are typically hundreds of gigabytes to terabytes, often over days or weeks.
- Extortion (T1657): Public naming on the leak site, email pressure campaigns to executives, and threats to publish data — sometimes including direct outreach to the victim's customers and partners to amplify pressure.
Exploitation Status
The specific initial access vector in the GE/Philips claims has not been publicly confirmed at the time of writing. Treat this as an active extortion campaign in progress: victim lists on Clop's site historically grow over subsequent weeks as downstream and supply-chain victims are identified. Organizations should assume that any internet-facing file transfer platform or enterprise application they operate is a candidate target for Clop and Clop-adjacent actors.
Detection & Response
Clop's quiet phase — staging and exfiltration — is where you win or lose. Once data is gone, the only remaining lever is incident response and legal/regulatory management. The detections below target the behaviors that matter: mass archive creation, abnormal outbound data volume, and webshell activity on file transfer infrastructure.
Sigma Rules
---
title: Mass Archive Creation by Compression Tools - Potential Data Staging
id: 3f8a2c41-7b6e-4d19-a2c5-9e1f4b8d6a02
status: experimental
description: Detects execution of archive utilities with password protection or multi-file compression consistent with ransomware/extortion data staging behavior observed in Clop-style exfiltration campaigns.
references:
- https://attack.mitre.org/techniques/T1560/001/
- https://www.bleepingcomputer.com/news/security/philips-and-ge-investigating-clop-ransomware-data-theft-claims/
author: Security Arsenal
date: 2026/02/09
tags:
- attack.collection
- attack.t1560.001
logsource:
category: process_creation
product: windows
detection:
selection_binary:
Image|endswith:
- '\rar.exe'
- '\7z.exe'
- '\7za.exe'
- '\7zr.exe'
- '\winzip64.exe'
selection_switches:
CommandLine|contains:
- ' a '
- ' -p'
- ' -hp'
- ' -m5'
- ' -v'
condition: selection_binary and selection_switches
falsepositives:
- Legitimate backup operations by IT administrators
- Software packaging workflows
level: high
---
title: Webshell or Script Dropped in Web Server Directory
id: 8c1d5e72-4a3b-4f68-b7d1-2e9a6c3f5b18
status: experimental
description: Detects creation of script files (ASPX, JSP, PHP) in web server or file-transfer application directories by non-standard processes, consistent with webshell deployment after exploitation of public-facing applications such as managed file transfer platforms.
references:
- https://attack.mitre.org/techniques/T1505/003/
- https://www.bleepingcomputer.com/news/security/philips-and-ge-investigating-clop-ransomware-data-theft-claims/
author: Security Arsenal
date: 2026/02/09
tags:
- attack.persistence
- attack.t1505.003
- attack.t1190
logsource:
category: file_event
product: windows
detection:
selection_paths:
TargetFilename|contains:
- '\inetpub\wwwroot\'
- '\webapps\'
- '\wwwroot\'
- '\htdocs\'
selection_extensions:
TargetFilename|endswith:
- '.aspx'
- '.ashx'
- '.jsp'
- '.jspx'
- '.php'
- '.war'
filter_legit:
Image|endswith:
- '\iisexpress.exe'
- '\devenv.exe'
- '\msdeploy.exe'
- '\w3wp.exe'
condition: selection_paths and selection_extensions and not filter_legit
falsepositives:
- Legitimate application deployments (filter by deployment tooling as needed)
level: high
---
title: Web Server Process Spawning Command Shell
id: 5b7e9a14-2d6c-4f81-a3b9-7c4e1d8f2a63
status: experimental
description: Detects web server or application server worker processes spawning command interpreters, a hallmark of webshell execution following exploitation of internet-facing applications including MFT platforms targeted by Clop.
references:
- https://attack.mitre.org/techniques/T1059/
- https://attack.mitre.org/techniques/T1505/003/
author: Security Arsenal
date: 2026/02/09
tags:
- attack.execution
- attack.t1059
- attack.t1505.003
logsource:
category: process_creation
product: windows
detection:
selection_parent:
ParentImage|endswith:
- '\w3wp.exe'
- '\httpd.exe'
- '\nginx.exe'
- '\tomcat9.exe'
- '\java.exe'
- '\javaw.exe'
selection_child:
Image|endswith:
- '\cmd.exe'
- '\powershell.exe'
- '\pwsh.exe'
- '\certutil.exe'
- '\bitsadmin.exe'
- '\curl.exe'
condition: selection_parent and selection_child
falsepositives:
- Rare; some legacy applications invoke shells from web workers - investigate all hits
level: critical
KQL Hunt (Microsoft Sentinel / Defender)
This query hunts for abnormally large outbound transfers from servers — the strongest signal in Clop-style exfiltration, since the gang steals data at volumes that stand out sharply against baseline traffic.
// Hunt: Anomalous high-volume outbound data transfers from servers (potential exfiltration)
// Baseline deviation approach - flags hosts sending significantly more data than their 14-day norm
let lookback = 14d;
let window = 1d;
let baseline = DeviceNetworkEvents
| where TimeGenerated between (ago(lookback) .. ago(window))
| where InitiatingProcessFolderPath !has_any ("windows\\system32", "program files\\windows defender")
| summarize BaselineBytes = avg(tolong(SentBytes)) by DeviceName, RemoteIP
| where BaselineBytes > 0;
DeviceNetworkEvents
| where TimeGenerated > ago(window)
| where RemoteIPType == "Public"
| summarize RecentBytes = sum(tolong(SentBytes)), DistinctDestinations = dcount(RemoteIP),
Processes = make_set(InitiatingProcessFileName, 10) by DeviceName, RemoteIP
| join kind=inner baseline on DeviceName, RemoteIP
| where RecentBytes > (BaselineBytes * 10) and RecentBytes > 500000000 // >10x baseline AND >500MB
| project DeviceName, RemoteIP, RecentBytes, BaselineBytes, DistinctDestinations, Processes
| order by RecentBytes desc;
// Hunt: Compression utility execution followed by network activity (staging then exfil)
DeviceProcessEvents
| where TimeGenerated > ago(7d)
| where FileName in~ ("rar.exe", "7z.exe", "7za.exe", "winzip64.exe")
| where ProcessCommandLine has_any (" -p", " -hp", " a ")
| summarize ArchiveOps = count(), FirstSeen = min(TimeGenerated), LastSeen = max(TimeGenerated),
Commands = make_set(ProcessCommandLine, 5) by DeviceName, AccountName
| where ArchiveOps >= 3
| order by ArchiveOps desc;
Velociraptor VQL
Use this artifact across your server fleet — particularly MFT appliances, web servers, and file servers — to identify staging activity and suspicious child processes of web services.
-- Hunt for data staging and webshell execution indicators on servers
-- Part 1: Suspicious child processes of web server processes
SELECT Pid, Ppid, Name, Exe, CommandLine, Username, CreateTime,
get_process(pid=Ppid).Name AS ParentName,
get_process(pid=Ppid).Exe AS ParentExe
FROM pslist()
WHERE ParentExe =~ '(w3wp|httpd|nginx|tomcat|java)'
AND Name =~ '(cmd|powershell|pwsh|certutil|bitsadmin|curl|rar|7z)'
-- Part 2: Recently created archives in staging-friendly directories
SELECT FullPath, Size, Mtime, Atime
FROM glob(globs=['C:/ProgramData/**/*.rar', 'C:/ProgramData/**/*.7z',
'C:/ProgramData/**/*.zip', 'C:/Windows/Temp/**/*.rar',
'C:/Windows/Temp/**/*.7z', 'C:/Users/Public/**/*.rar',
'C:/Users/Public/**/*.zip'])
WHERE Mtime > now() - 604800
AND Size > 10000000
ORDER BY Size DESC
Verification & Hardening Script
Run this PowerShell script on Windows servers hosting file transfer platforms or web applications to audit for common post-exploitation artifacts and validate exposure.
# Clop-Style Intrusion Exposure Audit - Run elevated on file transfer / web servers
# Author: Security Arsenal IR Team
$ReportPath = "C:\Temp\ExfilAudit_$(Get-Date -Format 'yyyyMMdd_HHmm').txt"
New-Item -Path C:\Temp -ItemType Directory -Force | Out-Null
# 1. Audit for suspicious archives staged in common locations in last 14 days
"=== [1] Recent large archives (potential staging) ===" | Out-File $ReportPath
$stagingPaths = @('C:\ProgramData','C:\Windows\Temp','C:\Users\Public','C:\Perflogs')
foreach ($p in $stagingPaths) {
if (Test-Path $p) {
Get-ChildItem -Path $p -Recurse -Include *.rar,*.7z,*.zip -ErrorAction SilentlyContinue |
Where-Object { $_.LastWriteTime -gt (Get-Date).AddDays(-14) -and $_.Length -gt 10MB } |
Select-Object FullName, Length, LastWriteTime | Format-Table -AutoSize | Out-File $ReportPath -Append
}
}
# 2. Check web root for recently modified script files (webshell indicator)
"=== [2] Recently modified web scripts (last 30 days) ===" | Out-File $ReportPath -Append
$webRoots = @('C:\inetpub\wwwroot','C:\xampp\htdocs')
foreach ($w in $webRoots) {
if (Test-Path $w) {
Get-ChildItem -Path $w -Recurse -Include *.aspx,*.ashx,*.jsp,*.php -ErrorAction SilentlyContinue |
Where-Object { $_.LastWriteTime -gt (Get-Date).AddDays(-30) } |
Select-Object FullName, LastWriteTime | Format-Table -AutoSize | Out-File $ReportPath -Append
}
}
# 3. Identify compression tool executions from process auditing / recent prefetch
"=== [3] Compression utility prefetch artifacts ===" | Out-File $ReportPath -Append
Get-ChildItem C:\Windows\Prefetch -ErrorAction SilentlyContinue |
Where-Object { $_.Name -match 'RAR|7Z|WINZIP' } |
Select-Object Name, LastWriteTime | Format-Table -AutoSize | Out-File $ReportPath -Append
# 4. List top outbound data talkers (last 24h, requires firewall/netflow - uses TCP session counts as proxy)
"=== [4] Active external connections from server processes ===" | Out-File $ReportPath -Append
Get-NetTCPConnection -State Established -ErrorAction SilentlyContinue |
Where-Object { $_.RemoteAddress -notmatch '^(10\.|172\.(1[6-9]|2[0-9]|3[01])\.|192\.168\.|127\.|::1|fe80)' } |
Group-Object OwningProcess | Sort-Object Count -Descending | Select-Object -First 15 |
ForEach-Object { $proc = Get-Process -Id $_.Name -ErrorAction SilentlyContinue
"{0} ({1}) - {2} connections" -f $proc.ProcessName, $_.Name, $_.Count } | Out-File $ReportPath -Append
# 5. Verify audit policies needed to catch this activity are enabled
"=== [5] Audit policy verification ===" | Out-File $ReportPath -Append
auditpol /get /subcategory:"Process Creation","File System" | Out-File $ReportPath -Append
if (-not ((auditpol /get /subcategory:"Process Creation") -match 'Success')) {
"WARNING: Process Creation auditing not enabled - enable and add command-line logging (Event 4688)" | Out-File $ReportPath -Append
}
Write-Host "Audit complete: $ReportPath" -ForegroundColor Green
Remediation & Defensive Actions
If you suspect involvement in this campaign — or simply want to harden against the next one — prioritize in this order:
Immediate (24-48 hours):
- Inventory internet-facing file transfer and web application platforms. Every MFT appliance, secure file sharing portal, and externally exposed enterprise app is a Clop-class target. If you don't have a current inventory, that's your first gap.
- Confirm patch status on all internet-facing appliances against the vendor's latest security advisories. Clop's time-to-exploit after disclosure has historically been days — patch SLAs for edge devices must match that reality, not a 30-day cycle.
- Pull egress data volume reports for the past 90 days on file servers, MFT platforms, and database hosts. Look for sustained outbound transfers to unfamiliar destinations — exfiltration in these campaigns often runs for weeks before the extortion note arrives.
- If you are in the GE or Philips supply chain, contact your vendor risk management team now. Determine what data of yours those organizations held, and monitor Clop's leak site channels and threat intel feeds for your organization's name.
Short term (this quarter):
- Deploy the detections above and baseline outbound transfer volumes per server. Exfiltration detection fails when there is no baseline to compare against.
- Enable command-line process auditing (Event 4688 with command line) on all servers if not already done — every rule in this post depends on it.
- Restrict egress from servers that have no business initiating large outbound transfers. MFT platforms should talk to known destinations only; deny-by-default egress rules on the perimeter firewall neutralize the exfil phase entirely.
- Encrypt and minimize data at rest on transfer platforms. Clop's leverage is the data itself — platforms configured to auto-purge transferred files dramatically reduce extortion impact.
Strategic:
- Rehearse the extortion scenario in your IR plan. The decision tree for "attacker claims they stole our data and is threatening publication" — legal counsel, breach notification obligations under HIPAA/state laws/contractual terms, negotiation posture, customer communication — must be decided before the incident, not during it.
- Treat third-party breach exposure as first-party risk. Both GE and Philips sit inside thousands of organizations' supply chains. Your data governance program should know exactly what sensitive data leaves your perimeter and lands in vendor systems.
The Clop operation has been running mass-exploitation-and-extortion campaigns for years, and the pattern is stable: an edge platform, a quiet dwell time, bulk theft, and public naming. Organizations that control egress, patch edge infrastructure aggressively, and baseline their data flows will find these intrusions in days rather than at the leak site.
Related Resources
Security Arsenal Incident Response Services AlertMonitor Platform Book a SOC Assessment incident-response Intel Hub
Is your security operations ready?
Get a free SOC assessment or see how AlertMonitor cuts through alert noise with automated triage.