WARD: Adversarially Robust Defense of Web Agents against Prompt Injections

Web agents can autonomously complete online tasks by interacting with websites, but their exposure to open web environments makes them vulnerable to prompt injection attacks embedded in HTML content or visual interfaces. Existing guard models still suffer from limited generalization to unseen...

Fedora 42 : php (2026-3a58db70ca)

The remote Fedora 42 host has a package installed that is affected by multiple vulnerabilities as referenced in the FEDORA-2026-3a58db70ca advisory. PHP version 8.4.21 07 May 2026 Core: Fixed bug GH-19983 GC assertion failure with fibers, generators and destructors. iliaal Fixed bug GH-21478...

Veritas: A Semantically Grounded Agentic Framework for Memory Corruption Vulnerability Detection in Binaries

Detecting memory corruption vulnerabilities in stripped binaries requires recovering object semantics, interprocedural propagation, and feasible triggers from low-level, lossy representations. Recent LLM-based approaches improve code understanding, but reliable detection still requires grounding ...

PT-2026-41118

Name of the Vulnerable Software and Affected Versions Amazon SageMaker Python SDK versions prior to 2.257.2 Amazon SageMaker Python SDK versions prior to 3.8.0 Description Missing integrity verification in the Triton inference handler allows a remote authenticated actor with S3 write access to th...

RHEL 9 : firefox (RHSA-2026:17688)

The remote Redhat Enterprise Linux 9 host has packages installed that are affected by multiple vulnerabilities as referenced in the RHSA-2026:17688 advisory. Mozilla Firefox is an open-source web browser, designed for standards compliance, performance, and portability. Security Fixes: firefox:...

Amazon SageMaker Python SDK 安全漏洞

Amazon SageMaker Python SDK is a development toolkit provided by Amazon, Inc., for building, training, and deploying machine learning models. Versions of the Amazon SageMaker Python SDK prior to v2.257.2 and v3.8.0 contained security vulnerabilities. These vulnerabilities stemmed from a lack of...

PT-2026-41117

Name of the Vulnerable Software and Affected Versions Amazon SageMaker Python SDK versions prior to 2.257.2 Amazon SageMaker Python SDK versions prior to 3.8.0 Description The ModelBuilder/Serve component stores sensitive information in cleartext. A remote authenticated actor with permissions to...

MCP Registry 安全漏洞

MCP Registry is an open-source MCP server application store developed by Model Context Protocol. Versions of MCP Registry prior to 1.7.9 contained security vulnerabilities. These vulnerabilities stemmed from OCI ownership verification skipping tag matching checks during HTTP 429 requests, which...

CVE-2026-36741

U-SPEED AC1200 Gigabit Wi-Fi Router Model: T18-21K V1.0 is vulnerable to Command Injection. The Network Time Protocol NTP configuration interface does not properly sanitize user-supplied input. An authenticated user with permission to configure NTP settings can inject arbitrary system commands...

MINI-GFW2-H346-4263

Bulletin has no description...

CVE-2026-41674

A flaw was found in xmldom and @xmldom/xmldom, a JavaScript library for parsing and serializing XML. This vulnerability allows an attacker to inject arbitrary XML markup into a document due to improper handling of DocumentType node fields during serialization. By crafting malicious input, an...

MINI-966M-7HHV-CHFP

Bulletin has no description...

CVE-2026-31217

The loadmodel function in the neuralmagictraining.py script of the optimate project in commit a6d302f912b481c94370811af6b11402f51d377f 2024-07-21 allows arbitrary code execution. When a user supplies a directory path via the --model command-line argument, the function reads a module.py file from...

MINI-86CM-2GR9-R3GM

Bulletin has no description...

Microsoft's MDASH AI System Finds 16 Windows Flaws Fixed in Patch Tuesday

Microsoft has unveiled a new multi-model artificial intelligence AI-driven system called MDASH to facilitate vulnerability discovery and remediation at scale, adding that it's being tested by some customers as part of a limited private preview. MDASH, short for m ulti-mod el a gentic s canning h...

MINI-M8VH-8W2F-6R2J

Bulletin has no description...

MINI-X6VM-CC2P-F3CC

Bulletin has no description...

MINI-Q6XX-M23M-74HW

Bulletin has no description...

MINI-5HP4-9HCM-C448

Bulletin has no description...

OpenAI’s GPT-5.5 is as Good as Mythos at Finding Security Vulnerabilities

The UK's AI Security Institute evaluated GPT-5.5's ability to find security vulnerabilities, and found that it is comparable to Claude Mythos. Note that the OpenAI model is generally available. Here is the Institute's evaluation of Mythos. And here is an analysis of a smaller, cheaper model. It...