Learning the Wrong Lessons: Syntactic-Domain Spurious Correlations in Language Models

Whitepaper from researchers at MIT, Northeastern University, and Meta. For an LLM to correctly respond to an instruction it must understand both the semantics and the domain i.e., subject area of a given task-instruction pair. However, syntax can also convey implicit information Recent work shows...

Incorrect Authorization

Overview Affected versions of this package are vulnerable to Incorrect Authorization via the policy evaluation process when egress.toGroups.aws.securityGroupsIds references AWS security group IDs that do not exist or are not attached to any network interface. An attacker can gain broader outbound...

COGNITION: From Evaluation to Defense against Multimodal LLM CAPTCHA Solvers

This paper studies how multimodal large language models MLLMs undermine the security guarantees of visual CAPTCHA. We identify the attack surface where an adversary can cheaply automate CAPTCHA solving using off-the-shelf models. We evaluate 7 leading commercial and open-source MLLMs across 18...

Securing Large Language Models (LLMs) from Prompt Injection Attacks

Large Language Models LLMs are increasingly being deployed in real-world applications, but their flexibility exposes them to prompt injection attacks. These attacks leverage the model's instruction-following ability to make it perform malicious tasks. Recent work has proposed JATMO, a task-specif...

Large Language Models Cannot Reliably Detect Vulnerabilities in JavaScript: The First Systematic Benchmark and Evaluation

Researchers have proposed numerous methods to detect vulnerabilities in JavaScript, especially those assisted by Large Language Models LLMs. However, the actual capability of LLMs in JavaScript vulnerability detection remains questionable, necessitating systematic evaluation and comprehensive...

Red Teaming Large Reasoning Models

Large Reasoning Models LRMs have emerged as a powerful advancement in multi-step reasoning tasks, offering enhanced transparency and logical consistency through explicit chains of thought CoT. However, these models introduce novel safety and reliability risks, such as CoT-hijacking and...

Prompt Injection Through Poetry

In a new paper, "Adversarial Poetry as a Universal Single-Turn Jailbreak Mechanism in Large Language Models," researchers found that turning LLM prompts into poetry resulted in jailbreaking the models: Abstract : We present evidence that adversarial poetry functions as a universal single-turn...

Remote Code Execution (RCE)

Happy DOM is vulnerable to Remote Code Execution RCE. The vulnerability is due to the use of a non-isolated Node.js VM context with JavaScript evaluation enabled by default, which allows an attacker to run untrusted code that can escape the sandbox—potentially gaining access to process-level...

PT-2025-48268

The application contains an insecure 'redirectToUrl' mechanism that incorrectly processes the value of the 'redirectUrlParameter' parameter. The application interprets the entered string of characters as a Java expression, allowing an unauthenticated attacer to perform arbitrary code execution...

PT-2025-48264

Name of the Vulnerable Software and Affected Versions Apache CloudStack versions 4.18.0 through 4.20.1 Apache CloudStack versions 4.21.0 through 4.21.9 Description An improper control of code generation 'Code Injection' issue exists in Apache CloudStack, specifically within several APIs accessibl...

Constructing and Benchmarking: A Labeled Email Dataset for Text-Based Phishing and Spam Detection Framework

Phishing and spam emails remain a major cybersecurity threat, with attackers increasingly leveraging Large Language Models LLMs to craft highly deceptive content. This study presents a comprehensive email dataset containing phishing, spam, and legitimate messages, explicitly distinguishing betwee...

Zenitel TCIV-3+

RISK EVALUATION Successful exploitation of these vulnerabilities could result in arbitrary code execution or cause a denial-of-service condition. 2. RECOMMENDED PRACTICES CISA recommends users take defensive measures to minimize the risk of exploitation of these vulnerabilities, such as:...

Effective Command-Line Interface Fuzzing with Path-Aware Large Language Model Orchestration

Command-line interface CLI fuzzing tests programs by mutating both command-line options and input file contents, thus enabling discovery of vulnerabilities that only manifest under specific option-input combinations. Prior works of CLI fuzzing face the challenges of generating semantics-rich opti...

CVE-2025-63603

A command injection vulnerability exists in the MCP Data Science Server's reading-plus-ai/mcp-server-data-exploration 0.1.6 in the safeeval function src/mcpserverds/server.py:108. The function uses Python's exec to execute user-supplied scripts but fails to restrict the builtins dictionary in the...

CVE-2025-13035

The Code Snippets plugin for WordPress is vulnerable to PHP Code Injection in all versions up to, and including, 3.9.1. This is due to the plugin's use of extract on attacker-controlled shortcode attributes within the evaluateshortcodefromflatfile method, which can be used to overwrite the...

ICAM365 CCTV Camera Multiple Models

RISK EVALUATION Successful exploitation of these vulnerabilities could result in unauthorized exposure of camera video streams and camera configuration data. 2. RECOMMENDED PRACTICES CISA recommends users take defensive measures to minimize the risk of exploitation of these vulnerabilities, such...

Password Strength Analysis through Social Network Data Exposure: A Combined Approach Relying on Data Reconstruction and Generative Models

Although passwords remain the primary defense against unauthorized access, users often tend to use passwords that are easy to remember. This behavior significantly increases security risks, also due to the fact that traditional password strength evaluation methods are often inadequate. In this...

TencentOS Server 4: libxslt (TSSA-2025:0588)

The version of Tencent Linux installed on the remote TencentOS Server 4 host is prior to tested version. It is, therefore, affected by multiple vulnerabilities as referenced in the TSSA-2025:0588 advisory. Package updates are available for TencentOS Server 4 that fix the following vulnerabilities...

CVE-2025-13035 Code Snippets <= 3.9.1 - Authenticated (Contributor+) PHP Code Injection via extract() and PHP Filter Chains

The Code Snippets plugin for WordPress is vulnerable to PHP Code Injection in all versions up to, and including, 3.9.1. This is due to the plugin's use of extract on attacker-controlled shortcode attributes within the evaluateshortcodefromflatfile method, which can be used to overwrite the...

Small Language Models for Phishing Website Detection: Cost, Performance, and Privacy Trade-Offs

Phishing websites pose a major cybersecurity threat, exploiting unsuspecting users and causing significant financial and organisational harm. Traditional machine learning approaches for phishing detection often require extensive feature engineering, continuous retraining, and costly infrastructur...