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...

HIDBench: Benchmarking Large Language Models for Host-Based Intrusion Detection

Recent benchmark efforts have advanced the evaluation of large language models LLMs in cybersecurity, including tasks such as penetration testing and vulnerability identification. However, a critical cybersecurity task, namely intrusion detection from system logs, remains unexplored. In this work...

Not What You Asked For: Typographic Attacks in Household Robot Manipulation

Open-vocabulary embodied AI agents increasingly rely on vision-language models such as CLIP for object perception and task grounding. However, the shared embedding space that enables this flexibility introduces a structural vulnerability to typographic attacks, where printed text in a physical...

ContraFix: Agentic Vulnerability Repair Via Differential Runtime Evidence and Skill Reuse

Large language model LLM agents are increasingly used for automated vulnerability repair AVR, where repository-level reasoning enables them to inspect context and produce source-code patches. However, recent empirical results show that these agents still struggle with real-world vulnerabilities...

ADR: An Agentic Detection System for Enterprise Agentic AI Security

We present the Agentic AI Detection and Response ADR system, the first large-scale, production-proven enterprise framework for securing AI agents operating through the Model Context Protocol MCP. We identify three persistent challenges in this domain: 1 limited observability -- existing Endpoint...

How AI Hallucinations Are Creating Real Security Risks

AI hallucinations are introducing serious security risks into critical infrastructure decision-making by exploiting human trust through highly confident yet incorrect outputs. When an AI model lacks certainty, it doesn’t have a mechanism to recognize that. Instead, it generates the most probable...

Do Coding Agents Understand Least-Privilege Authorization?

As coding agents gain access to shells, repositories, and user files, least-privilege authorization becomes a prerequisite for safe deployment: an agent should receive enough authority to complete the task, without unnecessary authority that exposes sensitive surfaces.To study whether current...

DCVD: Dual-Channel Cross-Modal Fusion for Joint Vulnerability Detection and Localization

Software vulnerability detection plays a critical role in ensuring system security, where real-world auditing requires not only determining whether a function is vulnerable but also pinpointing the specific lines responsible. However, existing approaches either rely on a single information source...

Defense at AI speed: Microsoft’s new multi-model agentic security system tops leading industry benchmark

In this article 1. AI-powered vulnerability discovery at hyper-scale 2. Codename: MDASH—Microsoft Security’s new multi-model agentic scanning harness 3. Using codename MDASH for security research 4. The 5.12.2026 Patch Tuesday cohort 5. Two deep dives 1. CVE-2026-33827—Remote unauthenticated UAF ...

Do Androids Dream of Breaking the Game? Systematically Auditing AI Agent Benchmarks with BenchJack

Agent benchmarks have become the de facto measure of frontier AI competence, guiding model selection, investment, and deployment. However, reward hacking, where agents maximize a score without performing the intended task, emerges spontaneously in frontier models without overfitting. We argue tha...

SkillSafetyBench: Evaluating Agent Safety under Skill-Facing Attack Surfaces

Reusable skills are becoming a common interface for extending large language model agents, packaging procedural guidance with access to files, tools, memory, and execution environments. However, this modularity introduces attack surfaces that are largely missed by existing safety evaluations: eve...

LITMUS: Benchmarking Behavioral Jailbreaks of LLM Agents in Real OS Environments

The rapid proliferation of LLM-based autonomous agents in real operating system environments introduces a new category of safety risk beyond content safety: behavior jailbreak, where an adversary induces an agent to execute dangerous OS-level operations with irreversible consequences. Existing...

MonitoringBench: Semi-Automated Red-Teaming for Agent Monitoring

We introduce a red-teaming methodology that exposes harder-to-catch attacks for coding-agent monitors, suggesting that current practices may under-elicit attacks and overstate monitor performance. We identify three challenges with current red-teaming. First, mode collapse in attack generation,...

Benchmarking Large Language Models for IoC Recovery under Adversarial Code Obfuscation and Encryption

Software obfuscation and encryption present persistent challenges for program comprehension and security analysis, particularly when adversaries conceal Indicators of Compromise IoCs such as IP addresses within source code. While Large Language Models LLMs have recently demonstrated remarkable...

LCC-LLM: Leveraging Code-Centric Large Language Models for Malware Attribution

LLMs are increasingly explored for malware analysis; however, current LLM-based malware attribution remains limited by unsupported indicators and insufficient code-level grounding for identifying malicious and vulnerable code segments. To address these limitations, this research introduces LCC-LL...

Pen-Strategist: A Reasoning Framework for Penetration Testing Strategy Formation and Analysis

Cyber threats are rapidly increasing, expanding their impact from large-scale enterprises to government services and individual users, making robust security systems increasingly essential. However, a significant shortage of skilled cybersecurity professionals exacerbates this challenge. While...

ARGUS: Defending LLM Agents against Context-Aware Prompt Injection

The rise of Large Language Model LLM agents, augmented with tool use, skills, and external knowledge, has introduced new security risks. Among them, prompt injection attacks, where adversaries embed malicious instructions into the agent workflow, have emerged as the primary threat. However,...

VulKey: Automated Vulnerability Repair Guided by Domain-Specific Repair Patterns

The increasing prevalence of software vulnerabilities highlights the need for effective Automatic Vulnerability Repair AVR tools. While LLM-based approaches are promising, they struggle to incorporate structured security knowledge from sources like CWE and NVD. Current methods either use this...

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-2026-7510

The CVE-2026-7510 entry concerns OWAP DefectDojo up to 2.55.4, with an authorization bypass affecting the Benchmark/Engagement/Product/Survey functionality. The issue is reachable remotely and is supported by a public disclosure; upgrading to DefectDojo 2.56.0 addresses the vulnerability (patch e...