Ouanilo MEDEGAN 🇧🇯 retweeted
有人发现了一种非常离谱的 JavaScript 混淆技术：通过在代码里塞入大量生物武器、毒药和致死剂量相关的函数名，让 AI 因为安全限制而拒绝分析代码。
安全研究员 veritas 在分析浏览器指纹脚本时，发现一段来自 ServicePipe 相关域名的 JavaScript，里面充斥着极其诡异的函数名：
computeRicinLethalDosageForWaterSupply
calculateDeadlyLethalSuicideAspirinDosage
光看名字，仿佛在研究生化武器。但实际上，这段代码只是在收集浏览器指纹，包括 User-Agent、GPU、Canvas 等信息。
研究员怀疑，这些名字是开发者故意设置的陷阱，专门用来阻止 AI 辅助逆向工程。
于是他做了个实验：
• Astra：拒绝分析
• Astra + Daybreak Blue：仍然拒绝
• GLM-5.3：成功分析
这就有意思了。
传统代码混淆是通过复杂控制流、加密字符串、无意义变量名，让人类和工具难以理解代码。
而这种新思路，直接瞄准了大模型的安全机制。
代码本身可能完全无害，但只要在函数名里塞进足够多的危险词汇，部分 AI 就可能主动放弃分析。
更讽刺的是，模型越严格遵守某些安全规则，面对这种混淆时反而可能越容易受到干扰。
当然，目前还没有证据证明开发者一定是故意这样设计，也不能说明这种方法对所有模型都有效。
但它揭示了一个值得关注的趋势：随着越来越多的逆向工程依赖 AI，未来的代码混淆技术可能不再只针对编译器、反编译器和人类研究员，还会专门针对大模型的安全判断机制。
以前是让代码难以理解，现在甚至可以尝试让 AI 不敢理解。
On entendait dire que les métiers manuels seraient ceux que l'IA ne remplaceraient pas..
You can now book our humanoid cleaners on a recurring subscription, for up to 3 hours per cleaning at tau-robotics.com.
Video is first take, uncut, shown at 4× speed. See the reply for detailed footage at 1× speed.
Ouanilo MEDEGAN 🇧🇯 retweeted
🚨 La sécurité des sauvegardes ne doit pas être un angle mort.
Veeam corrige 5 vulnérabilités, dont une critique (CVE-2026-108142, CVSS 9,1) permettant de contourner l’autorisation sans authentification.
🔧 Mettez à jour vos solutions Veeam.
🔗 csirt.bj/multiples-vulnerabi…
Ouanilo MEDEGAN 🇧🇯 retweeted
Du 16 au 20 novembre 2026, Cotonou devient la capitale de l'innovation africaine.
Trois rendez-vous pendant la Benin Digital & Innovation Week : le Cyber Africa Forum (16–17 novembre), l'AfriLabs Annual Gathering (18–20 novembre) et le Village de l'Innovation (18–20 novembre).
Ouanilo MEDEGAN 🇧🇯 retweeted
Du 16 au 20 novembre 2026, #Cotonou devient la capitale de l'innovation africaine.
Trois rendez-vous pendant la Benin Digital & Innovation Week : @CAF_forum, @AfriLabs Annual Gathering et le Village de l'Innovation.
#BeninDIW #AAG2026 #CAF2026 #VillageInnovation #BDIW2026
Ouanilo MEDEGAN 🇧🇯 retweeted
Une vulnérabilité affectant PAN-OS permettrait à un administrateur authentifié de compromettre un pare-feu et de modifier les règles de sécurité du réseau.
🟠 Criticité : Élevée
📊 Score CVSS : 8,6/10
🔗 Alerte : csirt.bj/ecriture-arbitraire…
Ouanilo MEDEGAN 🇧🇯 retweeted
This is what modern warfare looks like now...
Taking out drones with big sticks
Writer: Ian
Ouanilo MEDEGAN 🇧🇯 retweeted
they’re calling it the black October lmao.
> node.js: security vulnerability.
> PHP: critical RCE vulnerabilities.
> tensorlake: npm package infected.
> liteLLM: malicious PyPI + Docker releases
> Railway: customer secrets exposed
> zeabur - infrastructure breached.
> OpenCode: AI API keys exposed.
> OVHcloud CLI: official binaries backdoored.
> RankMath PRO: malware inside official plugin.
> TeamUpdraft: customer data compromised.
> ConnectWise: cloud backup identities compromised.
> SubQuery: backdoored npm package.
> GhostAction: GitHub repos with malicious workflows (related to the broader Shai-Hulud/ChainDrop activity).
> eMudhra: certificate authority keys compromised (with downstream impact on Tailscale).
this some complete madness bro. 1/3 of the internet is fucked.
Ouanilo MEDEGAN 🇧🇯 retweeted
Bulletin d'alerte du bjCSIRT sur des vulnérabilités d'exécution de code à distance affectant PHP
csirt.bj/execution-de-code-a…
Ouanilo MEDEGAN 🇧🇯 retweeted
Bulletin d'alerte du bjCSIRT sur une vulnérabilité d'exécution de code à distance affectant GitLab CE/EE
csirt.bj/execution-de-code-a…
Ca vient plus vite que prévu.. Uber Bagarre ou Skynet. Et voir des gens comme Pliny Le Libérateur interagir avec cette publication, ne présage rien de bon😅. Chacun son générateur d'EMP.
Ouanilo MEDEGAN 🇧🇯 retweeted
AI is testing the limits of post-quantum cryptography.
An Anthropic AI model found an attack that halved HAWK's effective security in just 60 hours, prompting its withdrawal from NIST consideration.
Itan Barmes of QIZ Security explains why cryptographic agility now matters more than ever.
Read: thehackernews.com/expert-ins…
Je répète : pour moi l'IA est une plus grande "menace" pour la crypto de notre ère, que le calcul quantique qui finira par venir.
Herkes Claude’un 26 yıllık NSA şifresini yarım saatte çözmesini yapay zeka zaferi diye alkışlıyor ama kimse asıl olaya bakmıyor.
Satoshi Nakamoto’nun 2008’de Bitcoin’i kurarken NSA’in P-256 eğrisine güvenmeyip neden secp256k1 seçtiği dün gece yeniden masaya geldi.
Çünkü 30 dakikada çözülen şey sadece bir sayı değil, küresel internetin 26 yıllık paranoyası: 🧵👇
Readers added context they thought people might want to know
Claude’un NIST P-256 seed’lerinin 1997’de lrand48 + Unix timestamp ile üretildiğini bulması doğru. Ancak bu bir kırılma değil; eğriler zayıflamadı, backdoor ihtimali azaldı. Atlanan adaylar ve Satoshi bağlantısı spekülasyon.
Kaynaklar:
anthropic.com/research/ecdsa…
words.filippo.io/seeds-solved/
Ouanilo MEDEGAN 🇧🇯 retweeted
OMG… the Tesla Roadster just went upside down on a half pipe and landed!
Ouanilo MEDEGAN 🇧🇯 retweeted
OH MY GODDDDDDD
ANDRE FILI WITH POSSIBLY THE KO OF THE YEAR 💥
WHAT A BRUTAL HEADKICK #UFCVegas122