False Alarms, Real Damage: Adversarial Attacks Using LLM-Based Models on Text-Based Cyber Threat Intelligence Systems

Cyber Threat Intelligence CTI has emerged as a vital complementary approach that operates in the early phases of the cyber threat lifecycle. CTI involves collecting, processing, and analyzing threat data to provide a more accurate and rapid understanding of cyber threats. Due to the large volume ...

CVE-2025-53485

SetTranslationHandler.php does not validate that the user is an election admin, allowing any even unauthenticated user to change election-related translation text. While partially broken in newer MediaWiki versions, the check is still missing. This issue affects Mediawiki - SecurePoll extension:...

CVE-2025-53485

The MediaWiki SecurePoll extension is vulnerable to a missing authorization flaw in SetTranslationHandler.php . Because the software fails to verify that a user is an election administrator, any user—including unauthenticated attackers—can modify election-related translation text. This issue impa...

CVE-2025-53485

SetTranslationHandler.php does not validate that the user is an election admin, allowing any even unauthenticated user to change election-related translation text. While partially broken in newer MediaWiki versions, the check is still missing. This issue affects Mediawiki - SecurePoll extension:...

CVE-2025-53485: Missing Authorization

SetTranslationHandler.php does not validate that the user is an election admin, allowing any even unauthenticated user to change election-related translation text. While partially broken in newer MediaWiki versions, the check is still missing. This issue affects Mediawiki - SecurePoll extension:...

VLAI: a RoBERTa-Based Model for Automated Vulnerability Severity Classification

This paper presents VLAI, a transformer-based model that predicts software vulnerability severity levels directly from text descriptions. Built on RoBERTa, VLAI is fine-tuned on over 600,000 real-world vulnerabilities and achieves over 82% accuracy in predicting severity categories, enabling fast...

CVE-2025-49846

wire-ios is an iOS client for the Wire secure messaging application. From Wire iOS 3.111.1 to before 3.124.1, messages that were visible in the view port have been logged to the iOS system logs in clear text. Wire application logs created and managed by the application itself were not affected,...

CVE-2025-49846 wire-ios accidentally logs message contents

wire-ios is an iOS client for the Wire secure messaging application. From Wire iOS 3.111.1 to before 3.124.1, messages that were visible in the view port have been logged to the iOS system logs in clear text. Wire application logs created and managed by the application itself were not affected,...

CVE-2025-49846

Summary : CVE-2025-49846 affects the Wire iOS client. For Wire iOS versions 3.111.1–3.124.1, messages visible in the viewport were logged in clear text to the iOS system logs due to canOpenUrl() being called with an invalid URL, with access requiring physical possession of an unlocked device. The...

CVE-2025-49846 wire-ios accidentally logs message contents

wire-ios is an iOS client for the Wire secure messaging application. From Wire iOS 3.111.1 to before 3.124.1, messages that were visible in the view port have been logged to the iOS system logs in clear text. Wire application logs created and managed by the application itself were not affected,...

CVE-2025-1709

Several credentials for the local PostgreSQL database are stored in plain text partially base64 encoded...

CVE-2025-1709

Several credentials for the local PostgreSQL database are stored in plain text partially base64 encoded...

CVE-2025-1709 CVE-2025-1709

Several credentials for the local PostgreSQL database are stored in plain text partially base64 encoded...

CVE-2025-1709

CVE-2025-1709 concerns Endress+Hauser MEAC300-FNADE4: information disclosure caused by local PostgreSQL credentials stored in plaintext (partially base64 encoded). Several connected sources reiterate that credentials are exposed, impacting confidentiality. Root cause: credentials stored in plaint...

Flatboard Pro cross-site scripting vulnerability

Flatboard Pro is an open source forum system by Flatboard. A cross-site scripting vulnerability exists in Flatboard Pro versions prior to 3.2.2, which stems from insufficient validation of inputs to the footertext and announcement parameters in config.php, and could lead to a stored cross-site...

PT-2025-27812 · Wire · Wire

Name of the Vulnerable Software and Affected Versions: Wire iOS versions 3.111.1 through 3.124.1 Description: The issue concerns the logging of messages in clear text to the iOS system logs when they are visible in the view port. This occurs due to the canOpenUrl function being called with an...

CVE-2025-49846: Improper Output Neutralization for Logs

wire-ios is an iOS client for the Wire secure messaging application. From Wire iOS 3.111.1 to before 3.124.1, messages that were visible in the view port have been logged to the iOS system logs in clear text. Wire application logs created and managed by the application itself were not affected,...

PT-2025-27772

Name of the Vulnerable Software and Affected Versions: PostgreSQL affected versions not specified Description: The issue involves several credentials for the local PostgreSQL database being stored in plain text, with some partially base64 encoded. Recommendations: At the moment, there is no...

UBUNTU-CVE-2024-35164

The terminal emulator of Apache Guacamole 1.5.5 and older does not properly validate console codes received from servers via text-based protocols like SSH. If a malicious user has access to a text-based connection, a specially-crafted sequence of console codes could allow arbitrary code to be...

CVE-2024-35164

The terminal emulator of Apache Guacamole 1.5.5 and older does not properly validate console codes received from servers via text-based protocols like SSH. If a malicious user has access to a text-based connection, a specially-crafted sequence of console codes could allow arbitrary code to be...