Dark Reading

Faux ChatGPT, Claude API Packages Deliver JarkaStealer

Two Python packages posing as tools to integrate with popular chatbots and provide API access are actually delivering "JarkaStealer," an infostealer designed to target potentially thousands of victims.

'GoIssue' Cybercrime Tool Targets GitHub Developers for Bulk Credential Theft

A cybercrime tool called GoIssue is being sold for $700 on a forum. It helps cyberattackers steal email addresses from GitHub profiles to use for further attacks like malware delivery and data breaches.

Citrix Issues Patches for Zero-Day Recording Manager Bugs

Citrix has released patches for two vulnerabilities in its Virtual Apps and Desktop technology that could allow remote attackers to escalate privileges or execute arbitrary code on affected systems.

Fake Copyright Infringement Emails Spread Rhadamanthys

The emails are automated, and they look like they come from legal representatives of real companies. Many of these companies are in the technology, media, and entertainment industries.

OWASP Beefs up GenAI Security Advice Amid Growing Deepfakes

OWASP's guidance focuses on building infrastructure for authenticating human identity in video calls, creating processes for financial transactions, and developing incident-response plans.

APT36 Deploys ElizaRAT and ApoloStealer in Attacks on Indian Targets

Over the past year, APT36 has been observed using three different versions of ElizaRAT in separate campaigns targeting Indian entities, with the latest version using Google Drive for command-and-control communications.

Okta Fixes Auth Bypass Bug After 3-Month Lull

Okta fixed a bug that could let hackers bypass authentication by using long usernames or lengthy domain names. Cybercriminals could exploit this security hole to access Okta AD/LDAP delegated authentication with just a username.

'CrossBarking' Attack Exposes Opera Browser Users via Private APIs

Guardio researchers demonstrated how hackers could exploit private APIs in Opera browser, gaining the ability to manipulate settings, hijack accounts, disable security extensions, add malicious extensions, and more.

ChatGPT can be Manipulated Using Hexadecimal Encoding

This technique leverages the model's ability to process multiple input formats and its focus on step-by-step instructions. By encoding malicious instructions in hexadecimal format, the model is tricked into decoding and executing them.

Russia Targets Ukrainian Military via Spoofed Recruitment App

Researchers from Google's Threat Intelligence Group (TAG) and Mandiant have identified a campaign named UNC5812 that uses a fake version of the "Civil Defense" tool to drop malware and spread misinformation.

Defend Against Threats with Cyber Fusion

Cyware is the leading provider of cyber fusion solutions that power threat intelligence sharing , end-to-end automation and 360-degree threat response.

Trending Tags