Quality-Diversity Evolution for Discovering Diverse Vulnerabilities in LLM Safety

Current approaches to LLM adversarial testing suffer from coverage gaps: manual red-teaming does not scale, LLM-as-attacker methods exhibit mode collapse, and gradient-based approaches produce uninterpretable gibberish. We introduce a quality-diversity evolutionary framework that operates at the...

Security Bulletin: IBM Operations Analytics - Log Analysis is affected by denial of service (DoS) due to Apache Commons FileUpload

Summary Apache Commons FileUpload in WebSphere Application Server Liberty is used by IBM Operations Analytics - Log Analysis as part of the parse and process HTTP requests for handling file uploads. CVE-2023-24998. Vulnerability Details CVEID:CVE-2023-24998 DESCRIPTION: Apache Commons FileUpload...

CVE-2026-7786 Jinan USR IOT Technology Limited (PUSR) USR-W610 RS232/485 to Wi-Fi/Ethernet Converter Use of Hard-coded Credentials

Jinan USR IOT Technology Limited PUSR USR-W610 RS232/485 to Wi-Fi/Ethernet Converter device firmware contains plaintext administrative credentials embedded in the firmware image. These credentials can be extracted through firmware analysis and used to authenticate to device services...

CVE-2026-45555 Roslyn CodeLens MCP Server: Untrusted Roslyn Analyzer Execution via get_diagnostics Leads to Arbitrary Code Execution

Roslyn CodeLens MCP Server is a Roslyn-based MCP server providing semantic code intelligence for .NET codebases. From 0.0.9 to 1.17.0, the getdiagnostics MCP tool loads and executes all DiagnosticAnalyzer assemblies referenced by the target solution without any allowlist, signature check, or user...

What’s in the container? Analyzing vulnerabilities, risks and protection with Kaspersky Container Security and the KIRA AI assistant

Introduction Containerization using Docker has become firmly established in modern development standards, significantly increasing the speed and convenience of deploying various services. Developers often use ready-made Docker images, making only minimal changes. The largest repository of contain...

MAL-2026-5028 Malicious code in sorenson-webfonts (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector d45b3e803fc04f697e067f5dfbc9a9c37878d1b7faed2ad4aea69dd9bed25c32 [email protected] is a hollow package: index.js is a 2-line stub 'use strict'; module.exports = ;, author/description fields are empty, and th...

MAECO-Lite: Modular Ontology for Dynamic Malware Analysis

Capturing dynamic malware behavior in a practical but still semantically precise manner remains a significant challenge in cyber threat intelligence. While standards such as MAEC and STIX provide widely adopted vocabularies for describing malware artifacts and observations, they represent data wi...

How to Compare the Security of Code Written by Humans to LLM-Generated Code

Large language models LLMs are rapidly transforming how software is created and maintained. Comparing LLM-generated code against human-written standards is essential to determine whether these new tools uphold or erode the security baselines established by professional developers. Yet, we lack a...

Joern 4.0.551

Joern is the bug hunter's workbench. With this tool, you can uncover attack surface, sloppy coding practices, and variants of known vulnerabilities using an interactive code analysis shell. Joern supports C, C++, LLVM bitcode, x86 binaries via Ghidra, JVM bytecode via Soot, and Javascript...

USR-W610 信任管理问题漏洞

USR-W610 is an industrial-grade serial-to-Wi-Fi networking module developed by USR. The USR-W610 has a trust management vulnerability, which stems from the inclusion of plaintext management credentials in the firmware. This vulnerability could allow credentials to be extracted through firmware...

Stochastic Analysis of Cybersecurity Defense Strategies under Single Attack Scenario

This research presents a novel stochastic framework for proactive cybersecurity defense timing under a single attack scenario. The approach models the defense process as a continuous observation mechanism in which the defense instant and the subsequent observation slot follow independent...

Thou Shall Not Pass: Gatekeeping Outbound TLS Connections

Despite the widespread use of Transport Layer Security TLS, its security guarantees are frequently compromised by outdated versions and misconfigurations. To analyze this problem, we collected more than 50 million TLS handshakes over a two-week period at our research institution, Fondazione Bruno...

MAL-2026-5027 Malicious code in @neon-i18n/core-ui (npm)

--- -= Per source details. Do not edit below this line.=- Source: ossf-package-analysis dbdc5bd090d8e85771f77fa3a7a113e08fbfb31de54ae399ed92565bdac246df The OpenSSF Package Analysis project identified '@neon-i18n/core-ui' @ 99.99.99 npm as malicious. It is considered malicious because: - The...

Malicious code in @neon-i18n/core-ui (npm)

--- -= Per source details. Do not edit below this line.=- Source: ossf-package-analysis dbdc5bd090d8e85771f77fa3a7a113e08fbfb31de54ae399ed92565bdac246df The OpenSSF Package Analysis project identified '@neon-i18n/core-ui' @ 99.99.99 npm as malicious. It is considered malicious because: - The...

MAL-2026-4839 Malicious code in hellowornd (npm)

--- -= Per source details. Do not edit below this line.=- Source: ghsa-malware 0e9b9637d126bc60120f015b0af88898fae5cf613a015fd572ab74d2554e6d7f Any computer that has this package installed or running should be considered fully compromised. All secrets and keys stored on that computer should be...

Malicious code in justsaying-docs (npm)

--- -= Per source details. Do not edit below this line.=- Source: ossf-package-analysis e1728e1b0cb2ea174743b9e437b707c768bb8979ba6299fedabfd49ea8a7d8e2 The OpenSSF Package Analysis project identified 'justsaying-docs' @ 2.4.4 npm as malicious. It is considered malicious because: - The package...

Dissecting the Black Box: Circuit-Level Analysis of LLM Vulnerability Detection

Large language models LLMs can detect software vulnerabilities, but how do they actually identify vulnerable code? We address this question using mechanistic interpretability; analyzing the internal computations of a neural network to understand its reasoning process.Using Circuit Tracer on...

Investigating Detection and Obfuscation of Prompt Injection Attacks against Software Reverse Engineering AI Agents

Agentic software reverse engineering systems are vulnerable to prompt injection attacks placed into the source code of executable binary files. This research demonstrates defensive tactics for detecting the presences of prompt injection strings in the decompiler output of adversarial example...

rustfs 安全漏洞

RustFS is a high-performance object storage system developed by RustFS. Versions of RustFS prior to 1.0.0-beta.2 contained security vulnerabilities. These vulnerabilities stemmed from the fact that the management router did not perform authentication on performance analysis endpoints, which could...

Automatically Attacking Software Reverse Engineering AI Agents

Software tools for reverse engineering executable binary files, such as Ghidra, enable malware analysts to safely conduct robust static analysis without having access to original source code. Coupled with the analytic power of large language models LLM, agentic systems enabled with tools, such as...