Earth Lamia Develops Custom Arsenal to Target Multiple Industries

Trend™ Research has been tracking an active APT threat actor named Earth Lamia, targeting multiple industries in Brazil, India and Southeast Asia countries at least since 2023. The threat actor primarily exploits vulnerabilities in web applications to gain access to targeted organizations...

One Surrogate to Fool Them All: Universal, Transferable, and Targeted Adversarial Attacks with CLIP

Deep Neural Networks DNNs have achieved widespread success yet remain prone to adversarial attacks. Typically, such attacks either involve frequent queries to the target model or rely on surrogate models closely mirroring the target model -- often trained with subsets of the target model's traini...

CVE-2024-26138

The XWiki licensor application, which manages and enforce application licenses for paid extensions, includes the document Licenses.Code.LicenseJSON that provides information for admins regarding active licenses. This document is public and thus exposes this information publicly. The information...

CVE-2023-28655

A malicious user could leverage this vulnerability to escalate privileges or perform unauthorized actions in the context of the targeted privileged users...

CVE-2023-0040

Versions of Async HTTP Client prior to 1.13.2 are vulnerable to a form of targeted request manipulation called CRLF injection. This vulnerability was the result of insufficient validation of HTTP header field values before sending them to the network. Users are vulnerable if they pass untrusted...

CVE-2023-41332

Cilium is a networking, observability, and security solution with an eBPF-based dataplane. In Cilium clusters where Cilium's Layer 7 proxy has been disabled, creating workloads with policy.cilium.io/proxy-visibility annotations in Cilium = v1.13 or io.cilium.proxy-visibility annotations in Cilium...

CVE-2021-41176

Pterodactyl is an open-source game server management panel built with PHP 7, React, and Go. In affected versions of Pterodactyl a malicious user can trigger a user logout if a signed in user visits a malicious website that makes a request to the Panel's sign-out endpoint. This requires a targeted...

CAIN: Hijacking LLM-Humans Conversations Via a Two-Stage Malicious System Prompt Generation and Refining Framework

Large language models LLMs have advanced many applications, but are also known to be vulnerable to adversarial attacks. In this work, we introduce a novel security threat: hijacking AI-human conversations by manipulating LLMs' system prompts to produce malicious answers only to specific targeted...

Robust LLM Fingerprinting Via Domain-Specific Watermarks

As open-source language models OSMs grow more capable and are widely shared and finetuned, ensuring model provenance, i.e., identifying the origin of a given model instance, has become an increasingly important issue. At the same time, existing backdoor-based model fingerprinting techniques often...

PRUNE: a Patching Based Repair Framework for Certifiable Unlearning of Neural Networks

It is often desirable to remove a.k.a. unlearn a specific part of the training data from a trained neural network model. A typical application scenario is to protect the data holder's right to be forgotten, which has been promoted by many recent regulation rules. Existing unlearning methods invol...

FABLE: a Localized, Targeted Adversarial Attack on Weather Forecasting Models

Deep learning-based weather forecasting models have recently demonstrated significant performance improvements over gold-standard physics-based simulation tools. However, these models are vulnerable to adversarial attacks, which raises concerns about their trustworthiness. In this paper, we first...

New HTTPBot Botnet Launches 200+ Precision DDoS Attacks on Gaming and Tech Sectors

Cybersecurity researchers are calling attention to a new botnet malware called HTTPBot that has been used to primarily single out the gaming industry, as well as technology companies and educational institutions in China. "Over the past few months, it has expanded aggressively, continuously...

Nosy Layers, Noisy Fixes: Tackling DRAs in Federated Learning Systems Using Explainable AI

Federated Learning FL has emerged as a powerful paradigm for collaborative model training while keeping client data decentralized and private. However, it is vulnerable to Data Reconstruction Attacks DRA such as "LoKI" and "Robbing the Fed", where malicious models sent from the server to the clie...

GHSA-W9Q3-G4P5-5Q2R sudo-rs Allows Low Privilege Users to Enumerate Privileges of Others

Summary Users with limited sudo privileges e.g. execution of a single command can list sudo privileges of other users using the -U flag. This doesn't happen with the original sudo. PoC The initial test has been done in a container running Ubuntu 24.04 and installing oxidizr, running sudo-rs 0.2.2...

CVE-2025-46718

sudo-rs is a memory safe implementation of sudo and su written in Rust. Prior to version 0.2.6, users with limited sudo privileges e.g. execution of a single command can list sudo privileges of other users using the -U flag. This vulnerability allows users with limited sudo privileges to enumerat...

PowerSchool Paid Ransom, Now Hackers Target Teachers for More

PowerSchool paid ransom after a major data breach; now hackers are targeting teachers and schools with direct extortion…...

State of ransomware in 2025

Global ransomware trends and numbers With the International Anti-Ransomware Day just around the corner on May 12, Kaspersky explores the ever-changing ransomware threat landscape and its implications for cybersecurity. According to Kaspersky Security Network data, the number of ransomware...

CVE-2025-31200

A memory corruption issue was addressed with improved bounds checking. This issue is fixed in iOS 18.4.1 and iPadOS 18.4.1, macOS Sequoia 15.4.1, tvOS 18.4.1, visionOS 2.4.1, watchOS 11.5. Processing an audio stream in a maliciously crafted media file may result in code execution. Apple is aware ...

CVE-2025-31201

This issue was addressed by removing the vulnerable code. This issue is fixed in iOS 18.4.1 and iPadOS 18.4.1, macOS Sequoia 15.4.1, tvOS 18.4.1, visionOS 2.4.1. An attacker with arbitrary read and write capability may be able to bypass Pointer Authentication. Apple is aware of a report that this...

4.7 million customers’ data accidentally leaked to Google by Blue Shield of California

Blue Shield of California leaked the personal data of 4.7 million people to Google after a Google Analytics misconfiguration. The tech giant may have used this data for targeted advertising, according to Blue Shield, which is one of the largest health insurers in the US. In a data breach notice o...