SecReEvalBench: a Multi-Turned Security Resilience Evaluation Benchmark for Large Language Models

The increasing deployment of large language models in security-sensitive domains necessitates rigorous evaluation of their resilience against adversarial prompt-based attacks. While previous benchmarks have focused on security evaluations with limited and predefined attack domains, such as...

Adversarial Suffix Filtering: a Defense Pipeline for LLMs

Large Language Models LLMs are increasingly embedded in autonomous systems and public-facing environments, yet they remain susceptible to jailbreak vulnerabilities that may undermine their security and trustworthiness. Adversarial suffixes are considered to be the current state-of-the-art...

Red Teaming the Mind of the Machine: a Systematic Evaluation of Prompt Injection and Jailbreak Vulnerabilities in LLMs

Large Language Models LLMs are increasingly integrated into consumer and enterprise applications. Despite their capabilities, they remain susceptible to adversarial attacks such as prompt injection and jailbreaks that override alignment safeguards. This paper provides a systematic investigation o...

TokenProber: Jailbreaking Text-To-Image Models Via Fine-Grained Word Impact Analysis

Text-to-image T2I models have significantly advanced in producing high-quality images. However, such models have the ability to generate images containing not-safe-for-work NSFW content, such as pornography, violence, political content, and discrimination. To mitigate the risk of generating NSFW...

RLSA-2024:5402 Important: thunderbird security update

Mozilla Thunderbird is a standalone mail and newsgroup client. Security Fixes: Thunderbird: 115.14/128.1 mozilla: Fullscreen notification dialog can be obscured by document content CVE-2024-7518 mozilla: Out of bounds memory access in graphics shared memory handling CVE-2024-7519 mozilla: Type...

RLSA-2024:5391 Important: firefox security update

Mozilla Firefox is an open-source web browser, designed for standards compliance, performance, and portability. Security Fixes: Firefox: 115.14/128.1 ESR mozilla: Fullscreen notification dialog can be obscured by document content CVE-2024-7518 mozilla: Out of bounds memory access in graphics shar...

CVE-2025-20974

Summary: CVE-2025-20974 concerns Samsung’s PackageInstallerCN. Multiple connected sources confirm the issue is due to improper handling of insufficient permissions in PackageInstallerCN before version 15.0.11.0, enabling a local attacker to bypass user interaction for requested installations. Aff...

Ensure That the Password Validity Is Set Correctly

If a password is not changed for a long time, the password is vulnerable to brute force cracking, which compromises system security. If the password validity period is set too short, the password needs to be changed frequently, increasing management costs. In addition, users may fail to log in...

SAMSUNG PackageInstallerCN 安全漏洞

SAMSUNG PackageInstallerCN is a package installer from Samsung South Korea. A security vulnerability exists in SAMSUNG PackageInstallerCN prior to version 15.0.11.0, which stems from mishandling of insufficient privileges, and could allow a local attacker to bypass user interaction for installati...

RockyLinux 8 : thunderbird (RLSA-2024:5402)

The remote RockyLinux 8 host has packages installed that are affected by multiple vulnerabilities as referenced in the RLSA-2024:5402 advisory. Thunderbird: 115.14/128.1 mozilla: Fullscreen notification dialog can be obscured by document content CVE-2024-7518 mozilla: Out of bounds memory access ...

Good News for Script Kiddies? Evaluating Large Language Models for Automated Exploit Generation

Large Language Models LLMs have demonstrated remarkable capabilities in code-related tasks, raising concerns about their potential for automated exploit generation AEG. This paper presents the first systematic study on LLMs' effectiveness in AEG, evaluating both their cooperativeness and technica...

Attack and Defense Techniques in Large Language Models: a Survey and New Perspectives

Large Language Models LLMs have become central to numerous natural language processing tasks, but their vulnerabilities present significant security and ethical challenges. This systematic survey explores the evolving landscape of attack and defense techniques in LLMs. We classify attacks into...

Ai.Txt: a Domain-Specific Language for Guiding AI Interactions with the Internet

We introduce ai.txt, a novel domain-specific language DSL designed to explicitly regulate interactions between AI models, agents, and web content, addressing critical limitations of the widely adopted robots.txt standard. As AI increasingly engages with online materials for tasks such as training...

Enhanced SSO session may experience problem accessing network shares or group policy updates

You have configured configured Enhanced SSO for Citrix workspace app and you are either on Windows 10 or Windows 11 endpoints. When you launch a desktop session and try to access the domain shares you may encounter the below prompt for credentials - You can also repoduce this issue by accessing t...

Graph of Attacks: Improved Black-Box and Interpretable Jailbreaks for LLMs

The challenge of ensuring Large Language Models LLMs align with societal standards is of increasing interest, as these models are still prone to adversarial jailbreaks that bypass their safety mechanisms. Identifying these vulnerabilities is crucial for enhancing the robustness of LLMs against su...

How Private Is Your Attention? Bridging Privacy with In-Context Learning

In-context learning ICL-the ability of transformer-based models to perform new tasks from examples provided at inference time-has emerged as a hallmark of modern language models. While recent works have investigated the mechanisms underlying ICL, its feasibility under formal privacy constraints...

DualBreach: Efficient Dual-Jailbreaking Via Target-Driven Initialization and Multi-Target Optimization

Recent research has focused on exploring the vulnerabilities of Large Language Models LLMs, aiming to elicit harmful and/or sensitive content from LLMs. However, due to the insufficient research on dual-jailbreaking -- attacks targeting both LLMs and Guardrails, the effectiveness of existing...

GraphAttack: Exploiting Representational Blindspots in LLM Safety Mechanisms

Large Language Models LLMs have been equipped with safety mechanisms to prevent harmful outputs, but these guardrails can often be bypassed through "jailbreak" prompts. This paper introduces a novel graph-based approach to systematically generate jailbreak prompts through semantic transformations...

Exploit for Improper Access Control in Papercut Papercut_Mf

CVE-2023-27350 This PoC demonstrates how it’s possible to byp...

Mozilla Firefox 安全漏洞

Mozilla Firefox is an open source web browser from the Mozilla Foundation in the United States. A security vulnerability exists in versions of Mozilla Firefox prior to 137, which stems from the fact that when an AI chatbot is activated in one tab after using another tab, the document title of the...