Enhanced MLLM Black-Box Jailbreaking Attacks and Defenses

Multimodal large language models MLLMs comprise of both visual and textual modalities to process vision language tasks. However, MLLMs are vulnerable to security-related issues, such as jailbreak attacks that alter the model's input to induce unauthorized or harmful responses. The incorporation o...

REx86: A Local Large Language Model for Assisting in X86 Assembly Reverse Engineering

Reverse engineering RE of x86 binaries is indispensable for malware and firmware analysis, but remains slow due to stripped metadata and adversarial obfuscation. Large Language Models LLMs offer potential for improving RE efficiency through automated comprehension and commenting, but cloud-hosted...

CVE-2025-11915 HTTP Desynchronisation in Vertex AI for certain third-party models

Connection desynchronization between an HTTP proxy and the model backend. The fixes were rolled out for all proxies in front of impacted models by 2025-09-28. Users do not need to take any action...

CVE-2025-11915 HTTP Desynchronisation in Vertex AI for certain third-party models

Connection desynchronization between an HTTP proxy and the model backend. The fixes were rolled out for all proxies in front of impacted models by 2025-09-28. Users do not need to take any action...

CVE-2025-11915

CVE-2025-11915 involves a connection desynchronization issue between an HTTP proxy and the Vertex AI model backend. Root cause: synchronization mismatch across the proxy-backend boundary. Impact is described as affecting certain third-party models, with fixes rolled out for all proxies in front o...

TP-Link Patches Four Omada Gateway Flaws, Two Allow Remote Code Execution

TP-Link has released security updates to address four security flaws impacting Omada gateway devices, including two critical bugs that could result in arbitrary code execution. The vulnerabilities in question are listed below - CVE-2025-6541 CVSS score: 8.6 - An operating system command injection...

TP-Link Patches Four Omada Gateway Flaws, Two Allow Remote Code Execution

TP-Link has released security updates to address four security flaws impacting Omada gateway devices, including two critical bugs that could result in arbitrary code execution. The vulnerabilities in question are listed below - CVE-2025-6541 CVSS score: 8.6 - An operating system command injection...

vLLM Authentication Middleware Auth Bypass Info Disclosure

The vulnerability of the AuthenticationMiddleware class in the library for working with Large Language Models LLMs like vLLM is related to the use of hidden time channels for data transmission. Exploiting this vulnerability can allow an attacker to bypass security restrictions and disclose...

CVE-2025-60344

CVE-2025-60344 describes a path traversal vulnerability in the D-Link DSR series (DSR-150, DSR-150N, DSR-250N v1.09B32_WW). Unauthenticated remote attackers can manipulate file/directory path resolution (e.g., via “../”) due to insufficient input validation, potentially exposing sensitive system ...

HarmNet: A Framework for Adaptive Multi-Turn Jailbreak Attacks on Large Language Models

Large Language Models LLMs remain vulnerable to multi-turn jailbreak attacks. We introduce HarmNet, a modular framework comprising ThoughtNet, a hierarchical semantic network; a feedback-driven Simulator for iterative query refinement; and a Network Traverser for real-time adaptive attack...

Moxa Network Device Authentication Bypass

The vulnerability of Moxa network devices with microprogrammed software, such as EDR-G9010, EDR-8010, EDG-G1002-BP, TN-4900, NAT-102, NAT-108, and OnCell G4302-LTE4, is related to deficiencies in authentication mechanisms. Exploiting this vulnerability can allow unauthorized actors to gain...

CLASP: Cost-Optimized LLM-Based Agentic System for Phishing Detection

Phishing websites remain a significant cybersecurity threat, necessitating accurate and cost-effective detection mechanisms. In this paper, we present CLASP, a novel system that effectively identifies phishing websites by leveraging multiple intelligent agents, built using large language models...

Prompting the Priorities: A First Look at Evaluating LLMs for Vulnerability Triage and Prioritization

Security analysts face increasing pressure to triage large and complex vulnerability backlogs. Large Language Models LLMs offer a potential aid by automating parts of the interpretation process. We evaluate four models ChatGPT, Claude, Gemini, and DeepSeek across twelve prompting techniques to...

CVE-2025-5517

Heap-based Buffer Overflow vulnerability in ABB Terra AC wallbox UL40/80A, ABB Terra AC wallbox UL32A, ABB Terra AC wallbox MID/ CE -Terra AC MID, ABB Terra AC wallbox MID/ CE -Terra AC Juno CE, ABB Terra AC wallbox MID/ CE -Terra AC PTB, ABB Terra AC wallbox JP.This issue affects Terra AC wallbo...

CVE-2025-5517 Heap Memory Corruption Vulnerability

Heap-based Buffer Overflow vulnerability in ABB Terra AC wallbox UL40/80A, ABB Terra AC wallbox UL32A, ABB Terra AC wallbox MID/ CE -Terra AC MID, ABB Terra AC wallbox MID/ CE -Terra AC Juno CE, ABB Terra AC wallbox MID/ CE -Terra AC PTB, ABB Terra AC wallbox JP.This issue affects Terra AC wallbo...

CVE-2025-5517 Heap Memory Corruption Vulnerability

Heap-based Buffer Overflow vulnerability in ABB Terra AC wallbox UL40/80A, ABB Terra AC wallbox UL32A, ABB Terra AC wallbox MID/ CE -Terra AC MID, ABB Terra AC wallbox MID/ CE -Terra AC Juno CE, ABB Terra AC wallbox MID/ CE -Terra AC PTB, ABB Terra AC wallbox JP.This issue affects Terra AC wallbo...

CrossGuard: Safeguarding MLLMs against Joint-Modal Implicit Malicious Attacks

Multimodal Large Language Models MLLMs achieve strong reasoning and perception capabilities but are increasingly vulnerable to jailbreak attacks. While existing work focuses on explicit attacks, where malicious content resides in a single modality, recent studies reveal implicit attacks, in which...

Cisco IP Phones DoS

The vulnerability of the web interface of Microprogramming Software for Cisco IP Phones models 9800, 7800, 8800, and 8875 lies in the fact that the operation outputs data beyond the buffer in memory. Exploiting this vulnerability allows a malicious actor to cause service failure by sending a...

Multimodal Safety Is Asymmetric: Cross-Modal Exploits Unlock Black-Box MLLMs Jailbreaks

Multimodal large language models MLLMs have demonstrated significant utility across diverse real-world applications. But MLLMs remain vulnerable to jailbreaks, where adversarial inputs can collapse their safety constraints and trigger unethical responses. In this work, we investigate jailbreaks i...

Can Transformer Memory Be Corrupted? Investigating Cache-Side Vulnerabilities in Large Language Models

Even when prompts and parameters are secured, transformer language models remain vulnerable because their key-value KV cache during inference constitutes an overlooked attack surface. This paper introduces Malicious Token Injection MTI, a modular framework that systematically perturbs cached key...