Ghoul

[Ghoul @BashGhoul]

I do that :)

[Ahmad @TheAhmadOsman]

PRO TOP

Do not create Skills with markdown files, you should be creating Skills with deterministic scripts

[Z A D D Y @Zaddyzaddy]

In just the past 48 hours, 8 private HackerOne programs we have been invited to have paused submissions. I suspect programs aren’t just struggling with the influx of AI-generated reports. Budget constraints may also be playing a bigger role than we think.

[Digital EU 🇪🇺 @DigitalEU]

Your personal data matters. In Europe, online platforms must be more transparent about:

✅Your data collection
✅The way it’s stored
✅Data usage & deletion

Rules in action: https://t.co/Jw5V93KZXh

[ali @endingwithali]

october 2026 will go down as the most cybersecurity unaware month of all time

[gabsmashh @gabsmashh]

the DeepSeek order has some interesting implications beyond just banning the model itself: the language around third-party products developed using DeepSeek is very broad. we're going to see more friction between national security policy & the realities of AI dev very soon.

[Ghoul @BashGhoul]

@CristiVlad25 For certain tasks, it is.

[Tirana Report @Tirana_Report]

All of this is happening just so Edi Rama can stay Prime Minister of a corrupt regime. It is him against the people. #Albania

[Ghoul @BashGhoul]

@Bu_Kolthoum من الممكن

[Al Majalla @AlMajallaEN]

Is international law still capable of addressing today’s crises? In his new book, A World Without Law, Politics Without Principles, Ahmed Abdel Rabou explores this question in an article by Mohammed Magdy for #AlMajalla. https://t.co/7sIQ937vXJ

[gal౨ৎ @prtygal777]

such a precious moment

[Ghoul @BashGhoul]

ACAB

[Ahmad @ahm7nu]

اقلك المهاجرين هم الي سرقوا ونهبوا 🤣

هي الشرطة الفرنسية

[Israel Exposed عربي @IsraelExposedAr]

أطلقنا أداة مجانية مفتوحة المصدر تُدعى "Zero Breach"، يمكنها الكشف عما إذا كانت حملة تجسس إسرائيلية - كشف عنها مختبر "سيتيزن لاب" (Citizen Lab) مؤخراً - قد تركت أثراً على جهاز الكمبيوتر الخاص بك. في تلك الحملة، استُخدمت تقنيات إعلانية إسرائيلية لإصابة أجهزة الصحفيين عبر إعلانات عادية لا تتطلب أي تفاعل (أي لا تتطلب نقراً من المستخدم).

تتطلب الأداة تثبيت لغة البرمجة "بايثون" (Python) من الرابط: (https://t.co/TJtE7GQykZ) - تأكد من تفعيل خيار "Add python.exe to PATH" أثناء التثبيت. بعد ذلك، افتح واجهة "PowerShell" في الملف الذي حفظت فيه الأداة، ثم انسخ بداخله التالي:

python -X utf8 https://t.co/cbwJ3vUvLJ

تستغرق العملية حوالي 15 دقيقة. الأداة مخصصة للقراءة فقط؛ فهي لا تُجري أي تغييرات على جهازك، ولا تتطلب صلاحيات المسؤول (admin)، كما أنها لا تتصل بالإنترنت نهائياً. تقوم الأداة بفحص 338 مؤشراً لتحديد ما إذا كان أحد متصفحاتك قد تفاعل سابقاً مع أي من هذه الإعلانات الضارة (كما تتحقق أيضاً من وجود إصابات ببرمجية "Candiru").

يمكنك تحميل الأداة من هنا:
https://t.co/Hd5sYJ5rDI

كما يمكنك التحقق مما إذا كان متصفحك أو جهازك يحظر هذا النوع من الإعلانات تلقائياً - دون الحاجة لتحميل أي برامج - عبر الرابط التالي:
https://t.co/52JEIVxzUK

النتيجة المطلوبة هي ظهور علامة خضراء ✅. إذا لم تظهر، فعليك تعديل إعدادات متصفحك، أو تثبيت أداة لحظر الإعلانات، أو استخدام خدمة "DNS" توفر خاصية الحظر.

بالنسبة للهواتف: نظراً لأن عشرات الآلاف من التطبيقات تعرض هذه الإعلانات، ننصحك بإعداد خدمة "DNS" (مثل NextDNS أو AdGuard، أو استخدام قائمة حظر عبر جهاز التوجيه/الراوتر أو نظام Pi-hole)؛ ليقوم هاتفك بحظر هذه الإعلانات تلقائياً.

نوفر أيضاً قوائم حظر جاهزة يمكنك الاشتراك فيها؛ لذا ننصحك باستخدامها.

يُرجى الإبلاغ عن أي مشكلات تواجهك عبر قناتنا على "تيليجرام"

حفظكم الله.

[Ghoul @BashGhoul]

@ColdBrief @Xyu348 In Albania too. You're a Joke!

[OpenCode @opencode]

DeepSeek V4.1 Flash is now free in OpenCode

[Mohammed Fathy @secfathy]

كورس اختبار اختراق الذكاء الاصطناعي | AI Penetration Testing Course (Arabic) — الدرس الثاني
https://t.co/CrnOla6cwg

[Mohamed AbuMuslim (أبومُسْلِم) @m19o__]

The real problem I see since AI entered the security field is the lack of proper Computer Science fundamentals.

Today I see a lot of people talking about what AI can do and what it can't do, and when I read the replies defending their point of view I usually think the discussion is a waste of energy, because they are not going to listen anyway and nobody likes being told that they don't really understand what they are talking about, directly or indirectly.

If you want to judge something, you need to understand it first.

AI is not something that appeared yesterday. Research in AI has been going on since the 1950s, and if you want to understand the technology we have today, you need to understand how it evolved and how we reached this point.

For a field like SOC, AI is actually a perfect fit in many areas because at the end of the day you are dealing with Data.

A lot of SIEM solutions, EDRs and Monitoring and Detection products have been using Machine Learning for years. Do you really think there is a For loop going through every record one by one to give you insights and Alerts?

Classifier models have been around for a very long time and have been used heavily for exactly these kinds of problems, and I will come back to this in a second.

If you use an LLM to deal with a massive amount of Data, build the whole Workflow around it, then say AI failed because the Workflow failed, the problem may actually be your understanding of AI.

Not everything new and shiny is the right solution for every problem.

You can't use a fork to eat soup and then blame the fork. The problem is you.

Back to Classifiers and AI SOC.

When Jev came out as a Classifier model, it made a lot of things easier when it comes to speed and Data Classification for Monitoring. But it is still an engineering problem. You still need to know how to use it properly, where it is strong, where it is weak, and what you need to build around it to cover those weaknesses.

Now when you use a Classifier together with LLMs, build a proper Workflow and a solid Data Engineering Pipeline, then we can actually talk about whether AI can handle the process or not.

The problem I keep seeing is people treating everything as LLMs as if they are the same thing.

They are not.

And of course this is a simplified explanation, not the whole topic. I am just trying to make the point that before saying AI can or cannot do something, maybe first understand what you are actually using.

[Anthropic @AnthropicAI]

New on our Engineering blog: Creating in-house source control.

In May, we merged around 2,500 code changes per day. In September, we merged more than 30,000 per day. As a result, we became bottlenecked by developer infrastructure built for human (rather than agent) scale.

Here, we share how we built a new version control system—and fully migrated our monorepo from git—in five months: https://t.co/wzNkwHs7IN

[OpenAI Newsroom @OpenAINewsroom]

We discovered a sophisticated recording device that was hidden inside one of our offices. We used our models to understand the device and where it had come from, identifying that a specific employee had planted it a few days earlier. We referred the matter to the FBI, and we’re grateful to the agents and prosecutors who moved quickly to make an arrest. The FBI’s investigation is ongoing and we will continue to support their work.

At this time, we have no indications of this individual inappropriately accessing or exfiltrating model weights, source code, or customer data. The employee was terminated.

We take protecting our technology, data, and people seriously, and as a matter of national importance. Our security program is designed to protect against a wide range of threat actors, from insider risks to cybercriminals to nation states, and this case is an example of how AI can help address these sophisticated security threats.

[ᴅᴀɴɪᴇʟ ᴍɪᴇssʟᴇʀ 🛡️ @DanielMiessler]

Rich Hickey says LLMs are automation, not tools, since you make a request and the model acts on its own instead of you steering it like a hand tool.

https://t.co/cWpKZnNVxK

[المركز العربي @ArabCenter_ar]

📝 يتناول عزمي بشارة @AzmiBishara في هذا المقال أثر التعرّض المتواصل لخطاب إدارة الرئيـ...ـس الأمـ...ـيركي دونـ...ـالد تـ...ـرمـ...ـب وممارساته في القيم والأعراف الأخلاقية السائدة في المجال العمومي في الولايات المتحدة الأميركية وخارجها، انطلاقًا من التمييز بين الاعتبارات اللا-أخلاقية الخاصة بالسياسة وتبرير الخطاب والممارسة السياسية غير الأخلاقيين.

🔗 هذه الدراسة منشورة في دورية #سياسات_عربية: https://t.co/owJFgklI48