Two Views, One Truth: Spectral and Self-Supervised Features Fusion for Robust Speech Deepfake Detection

Recent advances in synthetic speech have made audio deepfakes increasingly realistic, posing significant security risks. Existing detection methods that rely on a single modality, either raw waveform embeddings or spectral based features, are vulnerable to non spoof disturbances and often overfit...

SDD: Self-Degraded Defense against Malicious Fine-Tuning

Open-source Large Language Models LLMs often employ safety alignment methods to resist harmful instructions. However, recent research shows that maliciously fine-tuning these LLMs on harmful data can easily bypass these safeguards. To counter this, we theoretically uncover why malicious fine-tuni...

CVE-2025-26397

SolarWinds Observability Self-Hosted is susceptible to Deserialization of Untrusted Data Local Privilege Escalation vulnerability. An attacker with low privileges can escalate privileges to run malicious files copied to a permission-protected folder. This vulnerability requires authentication fro...

CVE-2025-51863

Self Cross Site Scripting XSS vulnerability in ChatGPT Unli ChatGPTUnli.com thru 2025-05-26 allows attackers to execute arbitrary code via a crafted SVG file to the chat interface...

SolarWinds Observability Self-Hosted 代码问题漏洞

SolarWinds Observability Self-Hosted is an observation platform from US-based SolarWinds. A code issue vulnerability exists in SolarWinds Observability Self-Hosted that stems from untrustworthy data deserialization, which could lead to local elevation of privilege...

rsmail

Extension: RSMail! Version: Old 1.22.26, 1.22.27, 1.22.28 / New 1.22.29 Update details: Versions affected 1.19.20 through 1.22.28. Self XSS allows remote authenticated attackers to inject arbitrary web script or HTML via a crafted parameter. Fixed in 1.22.29 Update URL:...

hermes-agent

Hermes Agent ☤ The self-improving AI agent b...

LENS-DF: Deepfake Detection and Temporal Localization for Long-Form Noisy Speech

This study introduces LENS-DF, a novel and comprehensive recipe for training and evaluating audio deepfake detection and temporal localization under complicated and realistic audio conditions. The generation part of the recipe outputs audios from the input dataset with several critical...

CVE-2025-51858

The CVE-2025-51858 entry concerns a self XSS vulnerability in ChatPlayground.ai up to 2025-05-24, exploitable via crafted SVG content sent in the chat component. Affected software is ChatPlayground.ai (chat feature) with the root cause described as SVG content in chat triggering script execution ...

CVE-2025-51858

Self Cross-Site Scripting XSS vulnerability in ChatPlayground.ai through 2025-05-24, allows attackers to execute arbitrary code and gain sensitive information via a crafted SVG file contents sent through the chat component...

Exploit for CVE-2025-51863

CVE-2025-51863 Vulnerability description ChatGPTUtil is...

CVE-2025-49831

An attacker of Secrets Manager, Self-Hosted installations that route traffic from Secrets Manager to AWS through a misconfigured network device can reroute authentication requests to a malicious server under the attacker’s control. CyberArk believes there to be very few installations where this...

CVE-2025-49829

CVE-2025-49829 affects CyberArk Conjur: OSS prior to 1.22.1 and Secrets Manager, Self-Hosted prior to 13.5.1/13.6.1. Root cause is missing validations in Secrets Manager, Self-Hosted allowing authenticated attackers to inject resources into the database and bypass permission checks. Impacts inclu...

Important: Red Hat Security Advisory: Red Hat Developer Hub 1.5.3 release.

Red Hat Developer Hub 1.5.3 has been released. Red Hat Developer Hub RHDH is Red Hat's enterprise-grade, self-managed, customizable developer portal based on Backstage.io. RHDH is supported on OpenShift and other major Kubernetes clusters AKS, EKS, GKE. The core features of RHDH include a single...

microcode_ctl: From CVEorg collector

New Spectre-v2 attack classes have been discovered within CPU architectures that enable self-training exploitation of speculative execution within the same privilege domain. These novel techniques bypass existing hardware and software mitigations, including IBPB, eIBRS, and BHINO, by leveraging...

HASSLE: a Self-Supervised Learning Enhanced Hijacking Attack on Vertical Federated Learning

Vertical Federated Learning VFL enables an orchestrating active party to perform a machine learning task by cooperating with passive parties that provide additional task-related features for the same training data entities. While prior research has leveraged the privacy vulnerability of VFL to...

Several New Classes of Self-Orthogonal Minimal Linear Codes Violating the Ashikhmin-Barg Condition

Whitepaper called Several New Classes Of Self-Orthogonal Minimal Linear Codes Violating The Ashikhmin-Barg Condition...

Exploit for CVE-2025-49113

Roundcube RCE Lab CVE-2025-49113 !Open in GitHub Codespac...

tipc: fix NULL pointer dereference in tipc_mon_reinit_self()

...

PT-2025-31073

Name of the Vulnerable Software and Affected Versions Linux kernel affected versions not specified Description A use-after-free issue was identified in the rpl do srh inline function within the Linux kernel. The vulnerability occurs when the function fetches an IPv6 header ipv6 hdrskb and accesse...