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...

CVE-2025-1054: Improper Neutralization of Input During Web Page Generation

The UiCore Elements – Free Elementor widgets and templates plugin for WordPress is vulnerable to Stored Cross-Site Scripting via the UI Counter, UI Icon Box, UI Testimonial Slider, UI Testimonial Grid, and UI Testimonial Carousel widgets in all versions up to, and including, 1.0.16 due to...

Quantifying Source Speaker Leakage in One-To-One Voice Conversion

Using a multi-accented corpus of parallel utterances for use with commercial speech devices, we present a case study to show that it is possible to quantify a degree of confidence about a source speaker's identity in the case of one-to-one voice conversion. Following voice conversion using a...

SUSE CVE-2017-9333

OpenWebif 1.2.5 allows remote code execution via a URL to the CallOPKG function in the IpkgController class in plugin/controllers/ipkg.py, when the URL refers to an attacker-controlled web site with a Trojan horse package. This has security implications in cases where untrusted users can trigger...

BadApex: Backdoor Attack Based on Adaptive Optimization Mechanism of Black-Box Large Language Models

Previous insertion-based and paraphrase-based backdoors have achieved great success in attack efficacy, but they ignore the text quality and semantic consistency between poisoned and clean texts. Although recent studies introduce LLMs to generate poisoned texts and improve the stealthiness,...

Q-FAKER: Query-Free Hard Black-Box Attack Via Controlled Generation

Many adversarial attack approaches are proposed to verify the vulnerability of language models. However, they require numerous queries and the information on the target model. Even black-box attack methods also require the target model's output information. They are not applicable in real-world...

Bypassing Prompt Injection and Jailbreak Detection in LLM Guardrails

Large Language Models LLMs guardrail systems are designed to protect against prompt injection and jailbreak attacks. However, they remain vulnerable to evasion techniques. We demonstrate two approaches for bypassing LLM prompt injection and jailbreak detection systems via traditional character...

Token-Level Constraint Boundary Search for Jailbreaking Text-To-Image Models

Recent advancements in Text-to-Image T2I generation have significantly enhanced the realism and creativity of generated images. However, such powerful generative capabilities pose risks related to the production of inappropriate or harmful content. Existing defense mechanisms, including prompt...

CVE-2025-32169: Improper Neutralization of Input During Web Page Generation

Improper Neutralization of Input During Web Page Generation 'Cross-site Scripting' vulnerability in Suresh Prasad Showeblogin Social showeblogin-facebook-page-like-box allows DOM-Based XSS.This issue affects Showeblogin Social: from n/a through = 7.0...

CVE-2024-54767

An access control issue in the component /juisboxinfo.xml of AVM FRITZ!Box 7530 AX v7.59 allows attackers to obtain sensitive information without authentication. NOTE: this is disputed by the Supplier because it cannot be reproduced, and the issue report focuses on an unintended configuration wit...

Infinite loop

Overview Affected versions of this package are vulnerable to Infinite loop in the findBox function. An attacker can cause the application to hang indefinitely by supplying a malicious image. PoC js // mkdir 2.0.1 // cd 2.0.1/ // npm i [email protected] const imageSizeFromFile =...

GHSA-M5QC-5HW7-8VG7 image-size Denial of Service via Infinite Loop during Image Processing

Summary image-size is vulnerable to a Denial of Service vulnerability when processing specially crafted images. The issue occurs because of an infine loop in findBox when processing certain images with a box with size 0. Details If the first bytes of the input does not match any bytes in...

PT-2026-48232

Name of the Vulnerable Software and Affected Versions image-size versions 1.1.0 through 1.2.0 image-size versions 2.0.0 through 2.0.1 Description A denial of service issue exists when processing specially crafted images with zero-sized boxes. Remote attackers can cause an application hang by...

CVE-2025-31864: Improper Neutralization of Input During Web Page Generation

Improper Neutralization of Input During Web Page Generation 'Cross-site Scripting' vulnerability in Out the Box Beam me up Scotty beam-me-up-scotty allows Stored XSS.This issue affects Beam me up Scotty: from n/a through = 1.0.23...

PT-2026-48162

Name of the Vulnerable Software and Affected Versions GPAC MP4Box version 2.4 Description A NULL pointer dereference occurs in the ctts box write function within the isomedia/box code base.c file. This issue allows an attacker to cause a Denial of Service DoS by providing a specially crafted MP4...

CVE-2025-31450

Improper Neutralization of Input During Web Page Generation 'Cross-site Scripting' vulnerability in phantom.omaga Toggle Box toggle-box allows Stored XSS.This issue affects Toggle Box: from n/a through = 1.6...

CVE-2025-30830

Missing Authorization vulnerability in Hossni Mubarak Cool Author Box hm-cool-author-box-widget allows Exploiting Incorrectly Configured Access Control Security Levels.This issue affects Cool Author Box: from n/a through = 2.9.9...

CVE-2025-28885

Improper Neutralization of Input During Web Page Generation 'Cross-site Scripting' vulnerability in fiverraffiliates Fiverr.com Official Search Box fiverr-official-search-box allows Stored XSS.This issue affects Fiverr.com Official Search Box: from n/a through = 1.0.8...

WordPress Toggle Box plugin <= 1.6 - Cross Site Scripting (XSS) Vulnerability

Cross Site Scripting XSS Vulnerability discovered by muhammad yudha in WordPress Plugin Toggle Box versions = 1.6...

CVE-2025-31450

Improper Neutralization of Input During Web Page Generation 'Cross-site Scripting' vulnerability in phantom.omaga Toggle Box toggle-box allows Stored XSS.This issue affects Toggle Box: from n/a through = 1.6...