Strange PPT maldoc with low detection rates since 2022-02-02.
MD5: c0060c0741833af67121390922c44f91
PPT file>wscript.exe>powershell.exe>rundll32.exe
[+]Exec wscript when user moves mouse
[+]Wscript exec powershell
[+]PS download the XORed DLL.
[+]Exec it using rundll32.exe
Replying to @GenThreatLabs
Great research. @urlyzeio already catches this one: paste a suspicious Claude artifact or ChatGPT/Grok shared chat link, and we run it in an isolated browser, hit the copy button and show you the exact command it tries to plant, flagged as ClickFix.
app.urlyze.io/scan/db2da41e-…
Replying to @SquiblydooBlog
Still live today. We scanned the stage-2 URL hxxps://jdsw[.]alico[.]shop/Bin/ScreenConnect.ClientSetup.msi → Malicious, ScreenConnect named (self-hosted, off-vendor instance), while file analysis alone returned Clean. C2 + relay :8041 still up.
app.urlyze.io/scan/3aa28a3d-…
DOCGuard - Detect Maldocs in Seconds! retweeted
As far as I can tell, WESP's bet is this: most EDR blocking decisions are already table rules — this path, this signer, this parent, this token. Compile those into the kernel, push the rest to async telemetry, and you close both the CrowdStrike-style crash and the "agent hangs, box hangs" class in one move?
Replying to @Malwarehunterr
🔥ClickFix Internals, for the curious.🎣
🚀1,122 pages, 1,020 clipboard commands. Telegram descriptions as stager addresses, WebDAV mounted by pushd, gates served from smart contracts, a kit that prints "stop watching us :)" when it spots you.
🔎Ten weeks of reading, written up:
urlyze.io/blog/clickfix-at-s…
Somewhere, a threat actor has a UX designer.🎭🎨
Animated keyboards, countdown timers, a FAQ, "Privacy by Design", a 90-day access token, a fake gemini.google.com etc...
All that polish for one ask: Win+R, Ctrl+V, Enter.
10 ClickFix lures, caught by Urlyze 👇
app.urlyze.io/community/thre…
🚨 Fresh ClickFix lure impersonating OpenAI, caught the day it was registered
openai-credit[.]com was registered on Oct 2 via NICENIC. Urlyze flagged it as malicious the same day.
The kit is tiny (10 requests total) and runs a double impersonation:
→ Clones Cloudflare's "openai.com · Performing security verification" interstitial, down to the "Just a moment..." title
→ Clicking "Verify you are human" silently writes a shell command to the clipboard
→ Instructions: Win+R → Ctrl+V → Enter
→ A reassuring "Gateway Verification - Challenge Response ID" string tells the victim what they "will observe"
The irony: the site really is behind Cloudflare (AS13335, Cloudflare NS), yet the challenge is fake. A genuine Cloudflare challenge loads Cloudflare's challenge resources. This one loaded none, so the "gate" is just part of the page. That's how Urlyze told it apart from a real anti-bot wall.
It fits a clear trend: AI brands are now a go-to ClickFix lure. This week Huntress showed malicious Custom GPTs on chatgpt.com routing victims to fake Cloudflare CAPTCHAs. The credit-themed domain is likely aimed at developers and AI users hunting for API credits.
🔍 IOCs
Domain: openai-credit[.]com
Registrar: NICENIC International Group (created 2026-10-02)
Hosting: Cloudflare-proxied (don't block the edge IP)
TLS: Let's Encrypt YE1
DOM hash (pivot for sibling kits):
d8babece34fe49a18789b1e1b934f1c5cc63549ad0a8cee7b33b92b620a9db41
🛡️ Users: no real CAPTCHA ever asks you to open Run or a terminal.
Defenders: alert on powershell/mshta/cmd spawned by explorer.exe right after browser activity, and hunt the RunMRU registry key.
Full report 👇
app.urlyze.io/scan/0d070603-…
#ClickFix #OpenAI #Phishing #ThreatIntel #InfoSec
Replying to @Malwarehunterr
Replying to @TheM3gatr0n
Replying to @ShiinaaM @malwrhunterteam
DOCGuard - Detect Maldocs in Seconds! retweeted
Leaked Personal Data Turns e-Devlet #Phishing Into Personalized Financial Fraud
A phishing domain surfaced via @urlyzeio reveals a multi-stage operation impersonating Türkiye’s e-Devlet portal, going significantly beyond conventional credential harvesting. It demonstrates this by combining e-Devlet credential theft, identity-data lookups, and a fabricated legal case to ultimately demand payment.
Phishing domain: emaliyekontrol[.]com
Analysis
The operation begins by impersonating Türkiye's e-Devlet portal and several government services. Regardless of the selected service, the victim is eventually directed to a cloned e-Devlet authentication page requesting a T.C. identity number and e-Devlet password.
After credentials and identity information are submitted, the phishing infrastructure appears to query an external identity-data service. An exposed configuration file reveals API endpoints for T.C. identity and address lookups.
This provides the operators with a mechanism to validate or enrich information supplied by the victim before continuing the fraud.
Once the victim passes this stage, the phishing site presents a completely fabricated “Dava Dosyası Sorgulama” (Court Case Inquiry) page designed to resemble an official e-Devlet/UYAP workflow.
The victim is shown a personalized legal narrative containing elements such as:
→ Fake court case information and case numbers
→ Claims that the information is retrieved from UYAP
→ A fabricated warning regarding unpaid penalties and potential measures against bank accounts
→ A fake court-decision PDF populated with victim-specific information
→ A monetary penalty - in the observed instance, 50,000 TL
→ Bank account details presented as the destination for payment
→ Functionality for uploading the payment receipt
Victims are instructed to transfer 50,000 TL to the supplied IBAN and subsequently upload the payment receipt. IBANs, beneficiary names and phone numbers should be considered rotatable campaign artifacts rather than necessarily permanent infrastructure. Receiving accounts may also be compromised, rented, or controlled by money mules.
Attack chain:
e-Devlet impersonation → credential harvesting → identity/address lookup → fake court case → personalized PDF → financial pressure → IBAN payment
Rather than treating credential theft as the final objective, the operation uses validated personal information to make a victim-specific legal fraud scenario significantly more credible.
@TRCert @malwrhunterteam @haskologlu @0x6rss
Replying to @Malwarehunterr
Replying to @Malwarehunterr







































