Astra Linux - уязвимость в linux-6.12

In the Linux kernel, the following vulnerability has been resolved: bpf: Fix kmemleak warning for percpu hashmap Vlad Poenaru reported the following kmemleak issue: unreferenced object 0x606fd7c44ac8 size 32: backtrace crc 0: pcpuallocnoprof+0x730/0xeb0 bpfmapallocpercpu+0x69/0xc0...

Astra Linux - уязвимость в linux-6.12

In the Linux kernel, the following vulnerability has been resolved: bpf: Fix softlockup in arenamapfree on 64k page kernel On an aarch64 kernel with CONFIGPAGESIZE64KB=y, arenahtab tests cause a segmentation fault and soft lockup. The same failure is not observed with 4k pages on aarch64. It turn...

SOSBENCH: Benchmarking Safety Alignment on Scientific Knowledge

Large language models LLMs exhibit advancing capabilities in complex tasks, such as reasoning and graduate-level question answering, yet their resilience against misuse, particularly involving scientifically sophisticated risks, remains underexplored. Existing safety benchmarks typically focus...

CLSA-2025-1749562017 kernel: Fix of 24 CVEs

vsock: Keep the binding until socket destruction CVE-2025-21756 - mt76: fix use-after-free by removing a non-RCU wcid pointer CVE-2022-49328 - bpf, testrun: Fix alignment problem in bpfprogtestrunskb CVE-2022-49840 - ASoC: core: Fix use-after-free in sndsocexit CVE-2022-49842 - net: sched: Fix...

CLSA-2025-1749548218 Fix of 6 CVEs

CVE-url: https://ubuntu.com/security/CVE-2024-57896 - btrfs: flush delalloc workers queue before stopping cleaner kthread during unmount CVE-url: https://ubuntu.com/security/CVE-2024-56551 - drm/amdgpu: fix usage slab after free CVE-url: https://ubuntu.com/security/CVE-2021-47211 - ALSA: usb-audi...

Evaluation Empirique De La Sécurisation Et De L'Alignement De ChatGPT Et Gemini: Analyse Comparative Des Vulnérabilités Par Expérimentations De Jailbreaks

Large Language models LLMs are transforming digital usage, particularly in text generation, image creation, information retrieval and code development. ChatGPT, launched by OpenAI in November 2022, quickly became a reference, prompting the emergence of competitors such as Google's Gemini. However...

DAVSP: Safety Alignment for Large Vision-Language Models Via Deep Aligned Visual Safety Prompt

Large Vision-Language Models LVLMs have achieved impressive progress across various applications but remain vulnerable to malicious queries that exploit the visual modality. Existing alignment approaches typically fail to resist malicious queries while preserving utility on benign ones effectivel...

From Threat to Tool: Leveraging Refusal-Aware Injection Attacks for Safety Alignment

Safely aligning large language models LLMs often demands extensive human-labeled preference data, a process that's both costly and time-consuming. While synthetic data offers a promising alternative, current methods frequently rely on complex iterative prompting or auxiliary models. To address...

Joint-GCG: Unified Gradient-Based Poisoning Attacks on Retrieval-Augmented Generation Systems

Retrieval-Augmented Generation RAG systems enhance Large Language Models LLMs by retrieving relevant documents from external corpora before generating responses. This approach significantly expands LLM capabilities by leveraging vast, up-to-date external knowledge. However, this reliance on...

Why LLM Safety Guardrails Collapse after Fine-Tuning: a Similarity Analysis between Alignment and Fine-Tuning Datasets

Recent advancements in large language models LLMs have underscored their vulnerability to safety alignment jailbreaks, particularly when subjected to downstream fine-tuning. However, existing mitigation strategies primarily focus on reactively addressing jailbreak incidents after safety guardrail...

Watermarking Degrades Alignment in Language Models: Analysis and Mitigation

Watermarking techniques for large language models LLMs can significantly impact output quality, yet their effects on truthfulness, safety, and helpfulness remain critically underexamined. This paper presents a systematic analysis of how two popular watermarking approaches-Gumbel and KGW-affect...

Microsoft and CrowdStrike Launch Shared Threat Actor Glossary to Cut Attribution Confusion

Microsoft and CrowdStrike have announced that they are teaming up to align their individual threat actor taxonomies by publishing a new joint threat actor mapping. "By mapping where our knowledge of these actors align, we will provide security professionals with the ability to connect insights...

BitBypass: a New Direction in Jailbreaking Aligned Large Language Models with Bitstream Camouflage

The inherent risk of generating harmful and unsafe content by Large Language Models LLMs, has highlighted the need for their safety alignment. Various techniques like supervised fine-tuning, reinforcement learning from human feedback, and red-teaming were developed for ensuring the safety alignme...

Video Signature: In-Generation Watermarking for Latent Video Diffusion Models

The rapid development of Artificial Intelligence Generated Content AIGC has led to significant progress in video generation but also raises serious concerns about intellectual property protection and reliable content tracing. Watermarking is a widely adopted solution to this issue, but existing...

Safety Alignment Can Be Not Superficial with Explicit Safety Signals

Recent studies on the safety alignment of large language models LLMs have revealed that existing approaches often operate superficially, leaving models vulnerable to various adversarial attacks. Despite their significance, these studies generally fail to offer actionable solutions beyond data...

SafeCOMM: What about Safety Alignment in Fine-Tuned Telecom Large Language Models?

Fine-tuning large language models LLMs for telecom tasks and datasets is a common practice to adapt general-purpose models to the telecom domain. However, little attention has been paid to how this process may compromise model safety. Recent research has shown that even benign fine-tuning can...

Disrupting Vision-Language Model-Driven Navigation Services Via Adversarial Object Fusion

We present Adversarial Object Fusion AdvOF, a novel attack framework targeting vision-and-language navigation VLN agents in service-oriented environments by generating adversarial 3D objects. While foundational models like Large Language Models LLMs and Vision Language Models VLMs have enhanced...

GeneBreaker: Jailbreak Attacks against DNA Language Models with Pathogenicity Guidance

DNA, encoding genetic instructions for almost all living organisms, fuels groundbreaking advances in genomics and synthetic biology. Recently, DNA Foundation Models have achieved success in designing synthetic functional DNA sequences, even whole genomes, but their susceptibility to jailbreaking...

M3S-UPD: Efficient Multi-Stage Self-Supervised Learning for Fine-Grained Encrypted Traffic Classification with Unknown Pattern Discovery

The growing complexity of encrypted network traffic presents dual challenges for modern network management: accurate multiclass classification of known applications and reliable detection of unknown traffic patterns. Although deep learning models show promise in controlled environments, their...

Lifelong Safety Alignment for Language Models

LLMs have made impressive progress, but their growing capabilities also expose them to highly flexible jailbreaking attacks designed to bypass safety alignment. While many existing defenses focus on known types of attacks, it is more critical to prepare LLMs for unseen attacks that may arise duri...