memguard

MemGuard Software enclave for storage of sensitive information in memory. This package attempts to reduce the likelihood of sensitive data being exposed when in memory. It aims to support all major operating systems and is written in pure Go. Features Sensitive data is encrypted and authenticated...

Bounded Agents: Delegation Security for Multi-Agent AI Systems

LLM-based agents can act on behalf of a user to access cloud services, call tools, or invoke agents. At session start, the agent's permissions are set but remain static, and each request is evaluated independently, without considering prior actions. Within its permissions, an agent may act contra...

Prompt Injections for Defense

This seems to work: Researchers from Tracebit on Monday said they found that placing prompt injections alongside passwords, cryptographic keys, and other secrets stored on Amazon Web Services was often all that was needed to shut down attacks from AI hacking agents. The prompts direct the attacki...

CVE-2026-31709

In the Linux kernel SMB client (cifsacl), CVE-2026-31709 arises from insufficient validation of a server-provided DACL when rewriting security descriptors. The fix extends structural validation to ensure the DACL header, size, and per-ACE bounds are checked before any rewrite paths (replace_sids_...

Security update for zabbix

This update for zabbix fixes the following issues: CVE-2024-36469: Introduced clamping for mitigation of timing attacks. bsc1240676 CVE-2024-42325: Restricted access to user fields using user.get API method for users of User and Admin type, and restricted access to alert entities using alert.get...

TrapSuffix: Proactive Defense against Adversarial Suffixes in Jailbreaking

Suffix-based jailbreak attacks append an adversarial suffix, i.e., a short token sequence, to steer aligned LLMs into unsafe outputs. Since suffixes are free-form text, they admit endlessly many surface forms, making jailbreak mitigation difficult. Most existing defenses depend on passive detecti...

CVE-2025-68704

Jervis is a library for Job DSL plugin scripts and shared Jenkins pipeline libraries. Prior to 2.2, Jervis uses java.util.Random which is not cryptographically secure for timing attack mitigation. This vulnerability is fixed in 2.2...

Gcore Mitigates Record-Breaking 6 Tbps DDoS Attack

Luxembourg, Luxembourg, 14th October 2025, CyberNewsWire...

EUVD-2023-43662

Malicious code in bioql PyPI...

EUVD-2022-28281

Malicious code in bioql PyPI...

EUVD-2024-16344

Malicious code in bioql PyPI...

BeDKD: Backdoor Defense Based on Dynamic Knowledge Distillation and Directional Mapping Modulator

Although existing backdoor defenses have gained success in mitigating backdoor attacks, they still face substantial challenges. In particular, most of them rely on large amounts of clean data to weaken the backdoor mapping but generally struggle with residual trigger effects, resulting in...

PT-2025-24969 · Adobe · Experience Manager

Name of the Vulnerable Software and Affected Versions: Adobe Experience Manager versions 6.5.22 and earlier Description: A stored Cross-Site Scripting XSS issue affects the software, allowing a low-privileged attacker to inject malicious scripts into vulnerable form fields. When a victim browses ...

PT-2025-24952 · Adobe · Experience Manager

Name of the Vulnerable Software and Affected Versions: Adobe Experience Manager versions 6.5.22 and earlier Description: A stored Cross-Site Scripting XSS issue affects the software, allowing a low-privileged attacker to inject malicious scripts into vulnerable form fields. When a victim browses ...

PT-2025-24957 · Adobe · Experience Manager

Name of the Vulnerable Software and Affected Versions: Adobe Experience Manager versions 6.5.22 and earlier Description: A stored Cross-Site Scripting XSS issue affects the software, allowing a low-privileged attacker to inject malicious scripts into vulnerable form fields. When a victim browses ...

CVE-2023-26481

authentik is an open-source Identity Provider. Due to an insufficient access check, a recovery flow link that is created by an admin or sent via email by an admin can be used to set the password for any arbitrary user. This attack is only possible if a recovery flow exists, which has both an...

Robustness Via Referencing: Defending against Prompt Injection Attacks by Referencing the Executed Instruction

Large language models LLMs have demonstrated impressive performance and have come to dominate the field of natural language processing NLP across various tasks. However, due to their strong instruction-following capabilities and inability to distinguish between instructions and data content, LLMs...

Exploit for Improper Input Validation in Microsoft

CVE-2023-23397 Exploitation & Mitigation Demo 📌 Overview...

CVE-2025-31124

CVE-2025-31124 (Zitadel) describes a user enumeration flaw in the login flow caused by normalization of the username when the “Ignoring unknown usernames” setting is enabled. Although the UI prompts for a password and returns “Username or Password invalid” for non-existent users, the normalizatio...

PT-2025-10848

Name of the Vulnerable Software and Affected Versions Microsoft Windows versions prior to the patch released in March 2025. Description A security feature bypass vulnerability exists in Microsoft Management Console MMC. This vulnerability, also known as MSC EvilTwin CVE-2025-26633, allows an...