Piggybacking on Perception: Stealthy Concurrent Audio Prompt Injections against Multimodal LLM Agents

Large Language Model LLM-driven multimodal agents are increasingly deployed to execute autonomous tasks via continuous audio interaction. While this paradigm enhances interaction naturalness, it introduces a critical yet under-explored attack surface, as audio inputs inevitably contain...

Open Security Benchmark: Towards Autonomous Enterprise Cyber Defense

Enterprises are moving toward autonomous cyber defense: agentic AI that builds situational awareness of an organization's security state and reasons from it to assessments, decisions, and actions. This rests on a holistic view of the enterprise's security state, the continuous, cross-vendor pictu...

relay-secreport-bench

secreport-bench — the Security-Report Benchmark An evaluation...

GHSA-W879-237Q-WC7R vulnerabilities

Vulnerabilities for packages: crossplane-provider-azure-managedidentity, pulumi-language-yaml, pulumi-language-java, dagger, prometheus, vault-benchmark, zarf, cilium, flux-source-controller, docker-cli-buildx, go-discover, rootlesskit, k3s, istio, melange, kine, kubernetes, gomplate, containerd,...

Astra Linux – Vulnerability found in Linux 6.1, Linux 6.12

In the Linux kernel, the following vulnerability has been resolved: crypto: virtio – Add spinlock protection with virtqueue notification When a virtual machine boots with a virtio-crypto PCI device and a built-in backend, run the openssl benchmark command with multiple processes. For example:...

Siemens RUGGEDCOM RST2428P Missing Synchronization (CVE-2026-23229)

In the Linux kernel, the following vulnerability has been resolved: crypto: virtio - Add spinlock protection with virtqueue notification When VM boots with one virtio-crypto PCI device and builtin backend, run openssl benchmark command with multiple processes, such as openssl speed -evp aes-128-c...

Smarter Saboteurs, Better Fixers: Scaling and Security in Linear Multi-Agent Workflows

As LLM-based multi-agent systems MAS are deployed in the wild, the resilience of their collaboration structures against adversarial compromise becomes a critical safety concern. Attackers may leverage prompt-injection or jailbreaking to sabotage individual agents within MAS workflows, but the...

mem0 安全漏洞

mem0 is an open-source benchmark tool for efficient memory algorithms developed by Mem0. Versions of mem0 prior to 0.2.8 contain security vulnerabilities. These vulnerabilities stem from a lack of authorization verification, which may cause authenticated users with an API key to redirect all LLM...

NCMD: Benign-Anchored Feature Selection for Imbalanced Network Intrusion Detection

Feature selection is critical for network intrusion detection systems NIDS operating under high-dimensional, highly imbalanced traffic, as found in operational and defense networks. Traditional filter methods rank features using global statistics computed symmetrically across classes and thus fai...

CVE-2026-7510

A vulnerability was determined in OWAP DefectDojo up to 2.55.4. Affected by this vulnerability is an unknown functionality of the component Benchmark/Engagement/Product/Survey. Executing a manipulation can lead to authorization bypass. The attack can be executed remotely. The exploit has been...

CVE-2026-5970

A vulnerability was detected in FoundationAgents MetaGPT up to 0.8.1. This affects the function checksolution of the component HumanEvalBenchmark/MBPPBenchmark. Performing a manipulation results in code injection. The attack may be initiated remotely. The exploit is now public and may be used. Th...

ExploitGym AI Exploit Benchmark Tool

ExploitGym is a large-scale, realistic benchmark built from real-world vulnerabilities designed to evaluate AI agents' ability to develop exploits...

Domain-Conditioned Safety in Frontier Computer-Using Agents: A 793-Episode Browser Benchmark, a Coding-Domain Cross-Reference, and a Reproducibility Audit of Recent Red-Teaming

Recent computer-using-agent CUA red-teaming papers report prompt-injection attack success rates ASR of 42-98%, but these headline numbers cluster on retired models and on the most-vulnerable model in each paper's panel. We ask whether those techniques, reproduced as hand-crafted templates, still...

Towards Cybersecurity SuperIntelligence (CSI): What'S the Best Harness for Cybersecurity?

What is the best harness for cybersecurity AI? Cybersecurity systems are converging on a single execution scaffold per agent, an iterative shell loop driven by a Large Language Model LLM. However, scaffolds are not interchangeable, rarely interoperable, and no single scaffold dominates across all...

Relevance As a Vulnerability: How Web Retrieval Degrades Safety Alignment in LLM Agents

AI agents augment large language models with external tools such as web retrieval, enabling grounded and up-to-date responses. However, incorporating external content into the generation pipeline can weaken the safety alignment mechanisms that govern model outputs. Prior work shows that enabling...

SEC-Bench Pro: Can Language Models Solve Long-Horizon Software Security Tasks?

Large language models LLMs now support automated software security tasks, including vulnerability discovery and proof-of-concept PoC generation. Existing benchmarks do not faithfully evaluate LLMs in real-world bug hunting scenarios because they rely on fuzzing harnesses, target-specific...

CyberMaskQA: A Privacy-Aware Benchmark for Evaluating Large Language Models in Cybersecurity Question Answering

Large language models LLMs are increasingly applied to cybersecurity question answering QA for critical tasks such as incident response and vulnerability analysis. However, real-world operational contexts, including system logs and network configurations, inherently contain sensitive identifiers,...

Parser-Free Querying of Security Logs

Security analysts routinely query system logs to detect threats and investigate incidents, but each log source uses its own semi-structured format: logs are cheap to produce, but expensive to use. The standard approach, building per-source parsers to normalize logs into structured schemas, is...

UNAD+: An Explainable Hybrid Framework for Unknown Network Attack Detection

The detection of previously unseen network attacks remains a major challenge for intrusion detection systems. Although supervised learning methods often perform well on known attack classes, they are limited when new attack types are not represented in the training data. Unsupervised methods are...

Measuring Security without Fooling Ourselves: Why Benchmarking Agents Is Hard

The benchmarks used to evaluate AI agents in security-critical roles suffer from crucial weaknesses. Building on recent empirical evidence, we characterize three core challenges that undermine security evaluations: benchmark vulnerabilities, temporal staleness, and runtime uncertainty. We then...