نبض سایبری اخبار امنیت سایبری، با ذکر منبع
EN
Microsoft Security Blog

Detecting sleeper agents hiding in AI model neurons

Microsoft Research has released new threat research detailing how a capable agent model could be turned into a sleeper agent by editing fewer than ten neurons to inject a short payload; Microsoft Research has also found that this threat is detectable by frontier agents, which can identify anomalous weight edits. The post Detecting sleeper agents hiding in AI model neurons appeared first on Microsoft Security Blog.

بدافزار و آلودگی امنیت ابری امنیت هوش مصنوعی هوش تهدید هوش مصنوعی و یادگیری ماشین برنامه‌نویسی و زبان‌ها رایانش ابری دواپس و زیرساخت امنیت وب و نرم‌افزار پاسخ به رخداد و جرم‌یابی تست نفوذ
خواندن متن کامل در Microsoft Security Blog درخواست حذف

این صفحه فقط عنوان و خلاصه کوتاه خبر را نمایش می‌دهد. متن کامل در منبع اصلی منتشر شده است.

خبرهای مرتبط

آرشیو کامل ←
Help Net Security

Orca Security adds visibility into AI-generated code and agent risks

Orca Security has announced new capabilities to help organizations discover and secure AI-generated applications and autonomous agents across their cloud environments. The expansion builds on Orca’s AI Security offering, enabling security teams to uncover previously unknown AI applications, assess the risks associated with autonomous agents, and gain deeper visibility into AI-generated code. AI is fundamentally chang…

امنیت هوش مصنوعی
منبع
Ubuntu Security Notices

USN-9040-1: Linux kernel (Low Latency NVIDIA) vulnerabilities

It was discovered that some AMD processors did not properly perform Reverse Map Table (RMP) checks when the IOMMU accessed certain host buffers. A local attacker with hypervisor access could possibly use this to trigger an out-of-bounds condition and compromise the integrity of SEV-SNP guest memory. (CVE-2023-20585) Several security issues were discovered in the Linux kernel. An attacker could possibly use these to c…

آسیب‌پذیری و روز صفر امنیت اینترنت اشیا و صنعتی توسعه موبایل
منبع
CSO Online

What will it take to make VulnOps a reality?

Just over a year ago, before advanced AI models began discovering software vulnerabilities at a nearly unmanageable scale in the so-called vulnpocalypse, three industry experts pioneered the concept of “VulnOps.” VulnOps, as envisioned by Heather Adkins, Gadi Evron, and Bruce Schneier, is a significant phase in the evolution of AI-assisted cyber defense in which vulnerability research and remediation would become sca…

آسیب‌پذیری و روز صفر امنیت ابری امنیت هوش مصنوعی
منبع
NCSC Netherlands

NCSC-2026-0478 [1.00] [M/H] Kwetsbaarheden verholpen in N-able N-central

N-able heeft meerdere kwetsbaarheden verholpen in N-able N-central. Deze kwetsbaarheden treffen versies ouder dan 2026.4.1.26. Er meerdere kwetsbaarheden met betrekking tot XML External Entity (XXE), waardoor de verwerking van XML-data mogelijk op ongewenste wijze kan worden beïnvloed. Daarnaast zijn er kwetsbaarheden in de vorm van OGNL Injection en Stored Cross-Site Scripting (XSS), waarbij kwaadaardige code kan wo…

امنیت وب و نرم‌افزار
منبع