Linux kernel 安全漏洞

Linux kernel is the kernel used by Linux, the open source operating system of the Linux Foundation in the United States. A security vulnerability exists in the Linux kernel that stems from a race condition between cpu buffer writes and swaps, which could result in a warning being triggered...

Your Compiler Is Backdooring Your Model: Understanding and Exploiting Compilation Inconsistency Vulnerabilities in Deep Learning Compilers

Deep learning DL compilers are core infrastructure in modern DL systems, offering flexibility and scalability beyond vendor-specific libraries. This work uncovers a fundamental vulnerability in their design: can an official, unmodified compiler alter a model's semantics during compilation and...

Reasoning Introduces New Poisoning Attacks yet Makes Them More Complicated

Early research into data poisoning attacks against Large Language Models LLMs demonstrated the ease with which backdoors could be injected. More recent LLMs add step-by-step reasoning, expanding the attack surface to include the intermediate chain-of-thought CoT and its inherent trait of...

CVE-2025-57813

CVE-2025-57813 affects the traQ messenger (github.com/traPtitech/traQ). Before version 3.25.0, error handling during SQL queries can write sensitive data (e.g., OAuth tokens) to log files. An attacker with log access could trigger SQL errors to illicitly read recorded secrets. The issue is fixed ...

The Silent, Fileless Threat of VShell

The Silent, Fileless Threat of VShell By Sagar Bade · August 21, 2025 Introduction Linux environments are often seen as bastions of security, favored by developers, sysadmins, and security professionals for their stability, transparency, and resistance to malware. Compared to Windows, the attack...

Isolate Trigger: Detecting and Eradicating Evade-Adaptive Backdoors

All current detection of backdoor attacks on deep learning models fall under the category of a non essential featuresNEF, which focus on fighting against simple and efficient vertical class backdoor -- trigger is small, few and not overlapping with the source. Evade-adaptive backdoor EAB attacks...

BDFirewall: Towards Effective and Expeditiously Black-Box Backdoor Defense in MLaaS

In this paper, we endeavor to address the challenges of backdoor attacks countermeasures in black-box scenarios, thereby fortifying the security of inference under MLaaS. We first categorize backdoor triggers from a new perspective, i.e., their impact on the patched area, and divide them into:...

MAL-2025-6504 Malicious code in fastybird-triggers-module (PyPI)

--- -= Per source details. Do not edit below this line.=-...

Malicious code in fastybird-triggers-module (PyPI)

--- -= Per source details. Do not edit below this line.=-...

Thorium Platform Public Availability

Today, CISA, in partnership with Sandia National Laboratories, announced the public availability of Thoriumlink is external, a scalable and distributed platform for automated file analysis and result aggregation. Thorium enhances cybersecurity teams' capabilities by automating analysis workflows...

When LLMs Copy to Think: Uncovering Copy-Guided Attacks in Reasoning LLMs

Large Language Models LLMs have become integral to automated code analysis, enabling tasks such as vulnerability detection and code comprehension. However, their integration introduces novel attack surfaces. In this paper, we identify and investigate a new class of prompt-based attacks, termed...

Multi-Trigger Poisoning Amplifies Backdoor Vulnerabilities in LLMs

Recent studies have shown that Large Language Models LLMs are vulnerable to data poisoning attacks, where malicious training examples embed hidden behaviours triggered by specific input patterns. However, most existing works assume a phrase and focus on the attack's effectiveness, offering limite...

CVE-2025-53889 Directus missing permission checks for manual trigger Flows

Directus is a real-time API and App dashboard for managing SQL database content. Starting in version 9.12.0 and prior to version 11.9.0, Directus Flows with a manual trigger are not validating whether the user triggering the Flow has permissions to the items provided as payload to the Flow...

Security update for the Linux Kernel

The SUSE Linux Enterprise Micro 6.0 and 6.1 kernel was updated to receive various security bugfixes. The following security bugs were fixed: CVE-2024-57982: xfrm: state: fix out-of-bounds read during lookup bsc1237913. CVE-2024-58053: rxrpc: Fix handling of received connection abort bsc1238982...

SUSE-SU-2025:02254-1 Security update for the Linux Kernel

The SUSE Linux Enterprise 15 SP7 Azure kernel was updated to receive various security bugfixes. The following security bugs were fixed: - CVE-2023-52888: media: mediatek: vcodec: Only free buffer VA that is not NULL bsc1228557. - CVE-2024-49568: net/smc: check v2extoffset/eidcnt/ismgidcnt when...

Qualcomm Chipsets 资源管理错误漏洞

Qualcomm Chipsets are a family of chipsets from Qualcomm Incorporated USA. A resource management error vulnerability exists in Qualcomm Chipsets that stems from a memory corruption when handling private escape commands in event triggers...

Malicious code in jira-triggers (npm)

The package communicates with a domain associated with malicious activity. --- -= Per source details. Do not edit below this line.=- Source: ghsa-malware 7fe29e329352c213ad9899b9096425dab4709925d393b2b31c1e6144c2062478 Any computer that has this package installed or running should be considered...

MAL-2025-5424 Malicious code in jira-triggers (npm)

The package communicates with a domain associated with malicious activity. --- -= Per source details. Do not edit below this line.=- Source: ghsa-malware 7fe29e329352c213ad9899b9096425dab4709925d393b2b31c1e6144c2062478 Any computer that has this package installed or running should be considered...

CodeGuard: a Generalized and Stealthy Backdoor Watermarking for Generative Code Models

Generative code models GCMs significantly enhance development efficiency through automated code generation and code summarization. However, building and training these models require computational resources and time, necessitating effective digital copyright protection to prevent unauthorized lea...

Probe Before You Talk: Towards Black-Box Defense against Backdoor Unalignment for Large Language Models

Backdoor unalignment attacks against Large Language Models LLMs enable the stealthy compromise of safety alignment using a hidden trigger while evading normal safety auditing. These attacks pose significant threats to the applications of LLMs in the real-world Large Language Model as a Service...