This week's ThreatsDay roundup reads less like a collection of separate incidents and more like an indictment of a single defensive failure: implicit trust. An Android bulletin carrying roughly 200 patched vulnerabilities. Browser extensions granted sweeping permissions they never needed. Trusted services folded into social engineering chains. Legacy bugs still returning value to attackers. Exposed systems left exposed. A fake storefront ecosystem now numbering around 119,000 scam shops. Different headlines — same root cause. The path in was already open.
For defenders, the lesson is operational, not philosophical. Every one of these stories maps to a control you either have or don't: patch velocity, extension governance, DNS-layer filtering, and user resistance to paste-and-run lures. This post breaks down each theme and gives you the detections, hunts, and hardening steps to act on today.
Technical Analysis
1. Android Security Bulletin: ~200 Vulnerabilities
Google's latest Android security release addresses approximately 200 flaws across the platform — framework components, system libraries, media handling, kernel drivers, and chipset-specific (Qualcomm/MediaTek) components. Historically, bulletins of this size include a mix of elevation-of-privilege, remote code execution in media/parser components, and information disclosure bugs. The defender-relevant facts:
- Affected platforms: Android devices running security patch levels older than the current monthly bulletin. Devices from OEMs that lag the patch train (or have exited support entirely) remain exploitable long after disclosure.
- Attack surface: Media parsing components (Stagefright-class bugs), Bluetooth/Wi-Fi stack issues, and kernel driver EoPs are the categories that most often end up chained in commercial spyware and targeted mobile intrusion campaigns.
- Exploitation status: Bulletins of this scale routinely include at least a subset of flaws flagged as potentially under limited, targeted exploitation. Even absent confirmed in-the-wild use, the public diff of patched code gives sophisticated actors a roadmap for reverse-engineering working exploits within days.
Defensive takeaway: Mobile is a first-class attack surface. If your MDM isn't enforcing minimum security patch levels with compliance enforcement, you are accepting this risk silently.
2. Browser Extensions: Permission Overreach as an Attack Vector
The recurring pattern: an extension requests broad access — read and change all your data on all websites — delivers its advertised feature, and then monetizes or gets sold to a buyer who weaponizes the installed base. Once an extension holds activeTab, cookies, webRequest, or tabs permissions across all URLs, it can scrape session tokens, inject content, exfiltrate form data, and redirect traffic — all from inside the browser, below the visibility of most EDR.
Key risk mechanics:
- Ownership transfer: Legitimate extensions get acquired; the new owner pushes a malicious update to an existing, trusted install base. No new user action required.
- Update-channel abuse: Auto-update means today's benign extension is tomorrow's stealer without any reinstall event for your controls to catch.
- Token theft: Extensions with cookie access bypass MFA entirely by replaying session cookies.
3. Trusted Services in Social Engineering Chains
Attackers continue to chain legitimate services — cloud hosting, link shorteners, CAPTCHA gates, document-sharing platforms — into phishing and ClickFix-style flows. The ClickFix family of lures is particularly relevant: a fake error or CAPTCHA page instructs the user to paste and execute a command (PowerShell, mshta, or a Run-dialog invocation), converting the user into the execution mechanism. This defeats URL filtering and attachment sandboxing because nothing malicious is delivered — the victim types the malware invocation themselves.
Observable signature worth your SOC's attention: a browser process spawning a script interpreter. In a healthy environment, chrome.exe, msedge.exe, and firefox.exe essentially never parent powershell.exe, cmd.exe, mshta.exe, or wscript.exe. When they do, it is overwhelmingly either an extension gone rogue or a ClickFix-style social engineering execution.
4. 119,000 Scam Shops: Industrialized Fake E-Commerce
Takedown-resistant networks of fraudulent storefronts — now tracked at roughly 119,000 active sites — harvest payment card data and personal information at scale. They are promoted through malvertising, social ads, SEO poisoning, and direct phishing. For enterprises, the exposure is twofold: employees entering corporate cards on scam sites, and brand impersonation of your own storefront eroding customer trust. These networks rotate domains aggressively, which is why domain-age and DNS-reputation controls outperform static blocklists.
5. Old Bugs, Still Working; Exposed Systems, Still Exposed
The roundup's quieter theme: attackers keep pulling value from known-vulnerable, internet-facing systems that defenders never patched or decommissioned. There is no novelty to defend against here — only asset inventory, attack surface management, and patch SLA discipline. If your vulnerability management program measures anything, let it be time-to-remediation for internet-facing assets, because that is the metric attackers are exploiting.
Detection & Response
The highest-fidelity, lowest-noise detections from this week's themes center on browser-spawned execution (ClickFix/extension abuse), extension installation auditing, and outbound connections to newly registered domains (scam shops).
---
title: Browser Process Spawning Script Interpreter (ClickFix / Extension Abuse)
id: 8f3a1c92-4d7b-4e5a-9c21-2b6d8e4f7a10
status: experimental
description: Detects browsers spawning PowerShell, cmd, mshta, or wscript — a hallmark of ClickFix-style paste-and-run social engineering and malicious extension execution. Browsers should never parent script interpreters in normal operation.
references:
- https://attack.mitre.org/techniques/T1204/
- https://attack.mitre.org/techniques/T1059/001/
author: Security Arsenal
date: 2026/09/18
tags:
- attack.execution
- attack.t1204
- attack.t1059.001
logsource:
category: process_creation
product: windows
detection:
selection_parent:
ParentImage|endswith:
- '\chrome.exe'
- '\msedge.exe'
- '\firefox.exe'
- '\brave.exe'
- '\opera.exe'
selection_child:
Image|endswith:
- '\powershell.exe'
- '\pwsh.exe'
- '\cmd.exe'
- '\mshta.exe'
- '\wscript.exe'
- '\cscript.exe'
- '\rundll32.exe'
condition: selection_parent and selection_child
falsepositives:
- Rare legitimate browser-launched enterprise installers (verify parent command line)
level: high
---
title: Run Dialog or Explorer Spawned Encoded PowerShell (ClickFix Pattern)
id: 2c7e5b18-9a3f-4d61-8e44-5a9c1f3b6d02
status: experimental
description: Detects encoded or download-cradle PowerShell launched from explorer.exe or the Run dialog context, consistent with ClickFix lures that instruct victims to paste commands manually.
references:
- https://attack.mitre.org/techniques/T1059/001/
- https://attack.mitre.org/techniques/T1027/
author: Security Arsenal
date: 2026/09/18
tags:
- attack.execution
- attack.t1059.001
- attack.defense_evasion
- attack.t1027
logsource:
category: process_creation
product: windows
detection:
selection_parent:
ParentImage|endswith: '\explorer.exe'
selection_image:
Image|endswith:
- '\powershell.exe'
- '\pwsh.exe'
- '\mshta.exe'
selection_cli:
CommandLine|contains:
- ' -enc'
- ' -e '
- 'FromBase64String'
- 'IEX'
- 'Invoke-Expression'
- 'DownloadString'
- 'Start-BitsTransfer'
condition: selection_parent and selection_image and selection_cli
falsepositives:
- Admin quick-launch of encoded commands (should be rare; investigate each hit)
level: high
---
title: New Browser Extension Installed Outside Enterprise Policy
id: 5b9d2e71-6c4a-4f83-b217-8e3d7a5c9f14
status: experimental
description: Detects writes to Chrome/Edge extension preference and registry locations that may indicate a sideloaded or user-installed extension outside centrally managed allowlists. Baseline before deployment.
references:
- https://attack.mitre.org/techniques/T1176/
author: Security Arsenal
date: 2026/09/18
tags:
- attack.persistence
- attack.t1176
logsource:
category: registry_event
product: windows
detection:
selection:
TargetObject|contains:
- '\SOFTWARE\Policies\Google\Chrome\ExtensionInstallForcelist'
- '\SOFTWARE\Policies\Microsoft\Edge\ExtensionInstallForcelist'
- '\SOFTWARE\Wow6432Node\Google\Chrome\Extensions\'
- '\SOFTWARE\Wow6432Node\Microsoft\Edge\Extensions\'
condition: selection
falsepositives:
- Legitimate extension deployment via GPO or enterprise policy
- First-run browser setup
level: medium
// Hunt: Browser-spawned script interpreters (ClickFix / malicious extension execution)
// High-fidelity — browsers should never parent script engines. Tune only for known enterprise installer flows.
DeviceProcessEvents
| where TimeGenerated > ago(7d)
| where InitiatingProcessFileName in~ ("chrome.exe", "msedge.exe", "firefox.exe", "brave.exe", "opera.exe")
| where FileName in~ ("powershell.exe", "pwsh.exe", "cmd.exe", "mshta.exe", "wscript.exe", "cscript.exe", "rundll32.exe")
| project TimeGenerated, DeviceName, AccountName, InitiatingProcessFileName, InitiatingProcessCommandLine, FileName, ProcessCommandLine, ReportId
| order by TimeGenerated desc;
// Hunt: Outbound connections to newly registered domains (scam shop infrastructure)
// Assumes ingestion of proxy/firewall logs via CommonSecurityLog (CEF) or Defender network events.
DeviceNetworkEvents
| where TimeGenerated > ago(7d)
| where isnotempty(RemoteUrl)
| join kind=leftouter (
// If you maintain a domain-age enrichment table or use a TI feed, join here.
// Placeholder illustrates correlation against your scam-shop/TI indicator list.
externaldata(IndicatorValue:string)[h@"https://your-ti-feed.example/scam-shops.csv"] with (format="csv")
) on $left.RemoteUrl == $right.IndicatorValue
| project TimeGenerated, DeviceName, InitiatingProcessFileName, RemoteUrl, RemoteIP, ActionType
| order by TimeGenerated desc;
// Hunt: Syslog/CEF path for Linux-proxied or firewall-visible scam shop traffic
CommonSecurityLog
| where TimeGenerated > ago(7d)
| where DeviceVendor in ("Palo Alto Networks", "Zscaler", "Fortinet", "Cisco")
| where isnotempty(RequestURL) or isnotempty(DestinationHostName)
| where RequestURL contains "shop" or DestinationHostName matches regex @"[a-z0-9-]{20,}\.(top|shop|xyz|icu|store)$"
| summarize ConnectionCount = count() by SourceIP, DestinationHostName, RequestURL, DeviceAction
| where ConnectionCount > 3
| order by ConnectionCount desc;
-- Hunt: Browser processes with suspicious children + enumerate installed Chrome/Edge extensions
-- Combines live process inspection with on-disk extension artifact collection.
SELECT Pid, Ppid, Name, CommandLine, Exe, Username, CreateTime
FROM pslist()
WHERE Name =~ '(?i)(powershell|pwsh|cmd|mshta|wscript|cscript|rundll32)\.exe'
AND Ppid IN (
SELECT Pid FROM pslist()
WHERE Name =~ '(?i)(chrome|msedge|firefox|brave|opera)\.exe'
)
-- Separately, enumerate extension manifests across user profiles to inventory what's installed:
-- SELECT FullPath, Mtime, read_file(filename=FullPath, length=4096) AS ManifestHead
-- FROM glob(globs='C:/Users/*/AppData/Local/Google/Chrome/User Data/*/Extensions/*/*/manifest.json')
-- Review each manifest's "permissions" array for broad host access ("<all_urls>", "cookies", "webRequest").
# Browser Extension Audit + ClickFix Exposure Check
# Run via RMM/Intune/SCCM across the fleet. Outputs installed extensions with risky permissions.
$results = @()
$browserExtPaths = @(
"$env:LOCALAPPDATA\Google\Chrome\User Data",
"$env:LOCALAPPDATA\Microsoft\Edge\User Data"
)
# Enumerate all user profiles for multi-user systems
$profiles = Get-ChildItem "C:\Users" -Directory -ErrorAction SilentlyContinue
foreach ($profile in $profiles) {
foreach ($base in $browserExtPaths) {
$userBase = $base -replace [regex]::Escape($env:LOCALAPPDATA), "$($profile.FullName)\AppData\Local"
$extDirs = Get-ChildItem "$userBase\*\Extensions" -Directory -ErrorAction SilentlyContinue
foreach ($extDir in $extDirs) {
$manifests = Get-ChildItem $extDir.FullName -Recurse -Filter "manifest.json" -ErrorAction SilentlyContinue
foreach ($m in $manifests) {
try {
$manifest = Get-Content $m.FullName -Raw | ConvertFrom-Json
$perms = @($manifest.permissions) + @($manifest.host_permissions)
$risky = $perms | Where-Object { $_ -match 'all_urls|cookies|webRequest|tabs|clipboardRead|<all_urls>' }
$results += [PSCustomObject]@{
User = $profile.Name
ExtensionID = $extDir.Name
Name = $manifest.name
Version = $manifest.version
Permissions = ($perms -join ';')
RiskyPerms = ($risky -join ';')
RiskFlag = [bool]$risky
}
} catch { }
}
}
}
}
$results | Where-Object RiskFlag | Export-Csv "C:\ProgramData\ExtAudit_$(hostname)_$(Get-Date -Format yyyyMMdd).csv" -NoTypeInformation
Write-Output "Flagged $($($results | Where-Object RiskFlag).Count) extensions with high-risk permissions. Review CSV."
# Verify enterprise extension policy is enforced (Chrome + Edge)
$policies = @(
'HKLM:\SOFTWARE\Policies\Google\Chrome\ExtensionInstallAllowlist',
'HKLM:\SOFTWARE\Policies\Microsoft\Edge\ExtensionInstallAllowlist',
'HKLM:\SOFTWARE\Policies\Google\Chrome\ExtensionInstallBlocklist',
'HKLM:\SOFTWARE\Policies\Microsoft\Edge\ExtensionInstallBlocklist'
)
foreach ($p in $policies) {
if (Test-Path $p) { Write-Output "[OK] Policy present: $p" }
else { Write-Output "[MISSING] No policy at $p — extensions are user-controlled for this browser" }
}
# Check for suspicious RunMRU entries (ClickFix victims paste commands via Win+R)
$mru = Get-ItemProperty "HKCU:\Software\Microsoft\Windows\CurrentVersion\Explorer\RunMRU" -ErrorAction SilentlyContinue
if ($mru) {
$mru.PSObject.Properties | Where-Object { $_.Value -match 'powershell|mshta|cmd|curl|certutil|bitsadmin' } |
ForEach-Object { Write-Output "[ALERT] Suspicious RunMRU entry: $($_.Value)" }
}
# Android fleet: verify patch compliance via MDM (Microsoft Graph example — requires appropriate module/scope)
# Get-MgDeviceManagementManagedDevice | Where-Object { $_.OperatingSystem -eq 'Android' } |
# Select DeviceName, AndroidSecurityPatchLevel, ComplianceState |
# Where-Object { [datetime]$_.AndroidSecurityPatchLevel -lt (Get-Date).AddDays(-45) }
Remediation
Mobile (Android bulletin):
- Push the current monthly Android security patch level to all managed devices immediately. Enforce a maximum patch-age compliance policy in your MDM (recommend ≤30 days; hard fail at 45) with conditional access blocking non-compliant devices from corporate resources.
- Identify and inventory devices from OEMs that have exited security support. Replace or network-isolate them — an unpatchable phone with corporate email access is a standing foothold.
- Restrict sideloading (
unknown sources) via Android Enterprise policy, and audit for sideloaded APKs in your fleet.
Browser extension governance:
- Flip the default: deploy
ExtensionInstallBlocklist=*and an explicitExtensionInstallAllowlistfor Chrome and Edge via GPO/Intune. If full allowlisting isn't feasible, at minimum block the high-risk permission categories usingExtensionSettingspolicy withblocked_permissions(cookies, webRequest, clipboardRead,<all_urls>). - Run the audit script above fleet-wide. Remove or quarantine any extension with broad host permissions that isn't business-justified.
- Monitor for extension ownership changes — treat any auto-update to a previously vetted extension as a change-management event, not routine noise.
Social engineering (ClickFix-class):
- Deploy the browser-child-process Sigma rule and KQL hunt above. This single detection covers a disproportionate share of current lure tradecraft.
- Train users specifically on the pattern: no legitimate website will ever ask you to press Win+R and paste a command. Name the technique in awareness materials — vague "don't click suspicious links" training does not touch this vector.
- Consider Attack Surface Reduction rules blocking Office child processes and script obfuscation where operationally tolerable.
Scam shop exposure:
- Enforce DNS filtering with newly-registered-domain (NRD) blocking — domains under 30 days old should be blocked or warned by default. This is the single most effective control against rotating scam storefronts.
- Ingest scam-shop/threat-intel indicator feeds into your proxy and firewall, and alert on matches.
- Prohibit corporate card use outside an approved merchant list; use virtual card numbers where possible to contain card-harvesting fallout.
Exposed systems and legacy flaws:
- Run continuous external attack surface management; anything internet-facing gets a 7-day (critical) / 14-day (high) remediation SLA.
- If you can't patch it, isolate it or retire it. An exposed system that stays exposed is a decision, not an oversight — make it an explicit, risk-accepted, time-boxed one.
The connective tissue across all 23-plus stories this week is governance of trust: trust in extensions, trust in patch cadence, trust in what users will execute, trust in the perimeter you think you have. Close the implicit-trust paths and most of these headlines stop being your problem.
Related Resources
Security Arsenal Managed SOC Services AlertMonitor Platform Book a SOC Assessment soc-mdr Intel Hub
Is your security operations ready?
Get a free SOC assessment or see how AlertMonitor cuts through alert noise with automated triage.