#redteam_yap
Have you heard about MOTW (Mark of the Web)?
aight letโs cook
MOTW is basically a trust badge Windows slaps on files that come from the internet.
Think of it as Windows saying:
> โyoo hold up bro, youโre a guest here โ
Whatโs really happening?
* MOTW is implemented using NTFS Alternate Data Streams (ADS)
* When a file is downloaded from the internet, Windows attaches metadata like:
- ZoneId
- Source URL
- Referrer info
- Most common case: ZoneId = 3 (Internet Zone)
there are 5 zones
ZoneId, Meaning
0 Local computer
1 Local intranet
2 Trusted sites
3 Internet (most common for downloads)
4 Restricted sites
Once that tag exists, Windows components start acting paranoid:
- ๐ Microsoft SmartScreen
- ๐ Microsoft Office (Protected View)
- ๐ Other security-aware apps
Result?
Pop-ups like:
> โAre you _sure_ you want to open this file?โ
> โThis file came from an untrusted source.โ
From an attackerโs perspectiveโฆ yeah, thatโs annoying ๐
๐ Attacker mindset: Okay, how do we blend in?
Since MOTW depends on ADS, the game becomes:
> Deliver the payload without inheriting ADS
Especially useful for multi-stage payloads, loaders, or initial access files.
and yup there are well-known, still-used ways to do this :)
This is gold for lateral movement and internal phishing ๐ฏ
Real Threat Actors Using These Techniques
Groups known to abuse MOTW bypass paths:
- TA505
- APT38
- APT29
@AfroSec
Have you heard about MOTW (Mark of the Web)?
aight letโs cook
MOTW is basically a trust badge Windows slaps on files that come from the internet.
Think of it as Windows saying:
> โyoo hold up bro, youโre a guest here โ
Whatโs really happening?
* MOTW is implemented using NTFS Alternate Data Streams (ADS)
* When a file is downloaded from the internet, Windows attaches metadata like:
- ZoneId
- Source URL
- Referrer info
- Most common case: ZoneId = 3 (Internet Zone)
there are 5 zones
ZoneId, Meaning
0 Local computer
1 Local intranet
2 Trusted sites
3 Internet (most common for downloads)
4 Restricted sites
Once that tag exists, Windows components start acting paranoid:
- ๐ Microsoft SmartScreen
- ๐ Microsoft Office (Protected View)
- ๐ Other security-aware apps
Result?
Pop-ups like:
> โAre you _sure_ you want to open this file?โ
> โThis file came from an untrusted source.โ
From an attackerโs perspectiveโฆ yeah, thatโs annoying ๐
๐ Attacker mindset: Okay, how do we blend in?
Since MOTW depends on ADS, the game becomes:
> Deliver the payload without inheriting ADS
Especially useful for multi-stage payloads, loaders, or initial access files.
and yup there are well-known, still-used ways to do this :)
1. MOTW Evasion Techniques (Still Wildly Relevant) Container / Disk Image Formats
Examples:
- .iso
- .vhd
- .vhdx
- .img
Why this works:
- When mounted via Windows Explorer, files inside the virtual disk do NOT inherit MOTW
- Payload comes out looking โlocalโ ๐
> Still abused heavily in real-world campaigns btw.
2. Physical Transfer / Internal Copy
- USB devices
- Copying from another internal machine
No browser โ no ADS โ no MOTW
Old-school, but effective.
3. Internal Email Attachments
- Payload archived (ZIP/RAR)
- Password-protected archive
- Sent from a compromised internal mailbox
> Even in modern Microsoft 365 environments (as of 2026):
> Internal emails do NOT apply MOTW by default
> Unless orgs explicitly enforce custom policies
This is gold for lateral movement and internal phishing ๐ฏ
Real Threat Actors Using These Techniques
Groups known to abuse MOTW bypass paths:
- TA505
- APT38
- APT29
@AfroSec
โ4๐ฅ2โก1
Forwarded from Tech Nerd (Tech Nerd)
โComparison is the thief of joyโ is advice for people already in motion โฆ not for those who havenโt started
@selfmadecoder
@selfmadecoder
๐ฅ14โค2โก1
whaaaat an actual fuck is this man ๐ข๐คฏ๐คฏ like those ai chatbot and agent companies should really test their agents especially their RAG pipeline
i was just chatting wiz one chatbot and suddenly i got an idea to test it and when i do boom this happened, it reveals its code base with snippets lol :)
sorry for the image quality btw
@AfroSec
i was just chatting wiz one chatbot and suddenly i got an idea to test it and when i do boom this happened, it reveals its code base with snippets lol :)
sorry for the image quality btw
@AfroSec
๐คฏ7๐2๐1
Forwarded from Sirack's Universe
Anthropic's Claude Code Security Wipes Billions Off Cybersecurity Stocks in a Single Afternoon | Awesome Agents
https://awesomeagents.ai/news/claude-code-security-cybersecurity-stocks-crash/
https://awesomeagents.ai/news/claude-code-security-cybersecurity-stocks-crash/
Awesome Agents
Anthropic's Claude Code Security Wipes Billions Off Cybersecurity Stocks in a Single Afternoon | Awesome Agents
Anthropic announced Claude Code Security, an AI tool that found 500+ vulnerabilities missed for decades in open-source code. Within hours, JFrog lost 25%, CrowdStrike dropped 8%, and the cybersecurity ETF hit its lowest since November 2023.
๐4๐ญ1
Forwarded from Florida๐ธ
Machine Learning Augmented Attacks
Recent threat reporting shows malicious actors are actively using AI as a force multiplier for offensive operations.
The interesting part isnโt that AI can generate code(i mean we already knew that)
The real shift is:
-Polymorphic Malware: rapid generation of variants to evade signature-based detection.
-Linguistic Smoothing: low skill actors are now bypassing "broken English" red flags in phishing.
-Condensed Attack Cycles: Reconnaissance and profiling that took days now take minutes.
๐ดFor Security & Red Teams
Focus less on what is generated,and more on how fast and how often:-
1.Content Velocity: abnormal speeds in phishing deployment.
2. LLM Wrappers: malicious use of open-source models in botnets.
3. Linguistic Patterns: Fluency โ legitimacy
4.Adaptive Payload Mutation: watch for rapid iteration patterns rather than static signatures.
AI is now part of both the exploit chain and the defense stack,the advantage wonโt belong to whoever โuses AIโ,it will belong to whoever understands how it changes execution speed.
Recent threat reporting shows malicious actors are actively using AI as a force multiplier for offensive operations.
The interesting part isnโt that AI can generate code(i mean we already knew that)
The real shift is:
-Polymorphic Malware: rapid generation of variants to evade signature-based detection.
-Linguistic Smoothing: low skill actors are now bypassing "broken English" red flags in phishing.
-Condensed Attack Cycles: Reconnaissance and profiling that took days now take minutes.
๐ดFor Security & Red Teams
Focus less on what is generated,and more on how fast and how often:-
1.Content Velocity: abnormal speeds in phishing deployment.
2. LLM Wrappers: malicious use of open-source models in botnets.
3. Linguistic Patterns: Fluency โ legitimacy
4.Adaptive Payload Mutation: watch for rapid iteration patterns rather than static signatures.
AI is now part of both the exploit chain and the defense stack,the advantage wonโt belong to whoever โuses AIโ,it will belong to whoever understands how it changes execution speed.
๐ฅ4๐2
This media is not supported in your browser
VIEW IN TELEGRAM
๐คฃ8๐ญ4๐1
Forwarded from Ge'ez Techยฎ แแแ แดแญ
#RandomThought
The Cyber sec Community here on telegram is not that active and known, so why dont we support each other and how many Enthusiast alu milewun lemasayet asbealewu so personal channel'm hone big channel sharing about cyber sec, ena if you want me to mention it here, send me your channel link @geez2012bot and i will check it ena if it looks helpful i will forward some posts from it, bezawu entewaweqalen.
tinish neger bihonm, let me support be aqme biye newu. Maqachewunm i will share some.
@geeztechgroup @geezsecurity #geeztech
The Cyber sec Community here on telegram is not that active and known, so why dont we support each other and how many Enthusiast alu milewun lemasayet asbealewu so personal channel'm hone big channel sharing about cyber sec, ena if you want me to mention it here, send me your channel link @geez2012bot and i will check it ena if it looks helpful i will forward some posts from it, bezawu entewaweqalen.
tinish neger bihonm, let me support be aqme biye newu. Maqachewunm i will share some.
@geeztechgroup @geezsecurity #geeztech
โค6
Forwarded from INSA Cyber Talent Center
แ แแแต แฒแ แ แ
แจแแ แจแ แจNathan Hailu Interview แแฌ แแณ 2:00 LT แ แแฅแณ แฅแแ แปแแ แแญ แญแ แฅแแ ๐ด LIVE TONIGHT โ 2:00 LT
Interview with Natan Hailu
Co-Founder @ GeezSecurity | HTB Certified Penetration Tester ๐
Cybersecurity โข Ethical Hacking โข Bug Hunting
๐ t.me/insactc
Donโt miss it ๐
#LiveTonight #CyberSecurity
Interview with Natan Hailu
Co-Founder @ GeezSecurity | HTB Certified Penetration Tester ๐
Cybersecurity โข Ethical Hacking โข Bug Hunting
๐ t.me/insactc
Donโt miss it ๐
#LiveTonight #CyberSecurity
๐4