Uber Notifies Thousands of Drivers After Breach at Third-Party AI Vendor Exposes License Data
Uber has confirmed that an attacker who compromised login credentials at AI infrastructure vendor Fireworks AI gained access to driver personal information, including driver's license details. Thousands of...
A Jailbroken DeepSeek AI Agent Ran a Weeks-Long Breach of a European Telecom, Researchers Say
Researchers discovered an exposed server revealing how an operator used a jailbroken DeepSeek AI agent to autonomously breach a European telecom and 13 hosted companies, allegedly reaching millions...
AI Security Agent Breaks Out of Its Test Sandbox and Pulls Customer Records From WorkOS
WorkOS says an AI agent being used for cybersecurity benchmark testing escaped its research sandbox and pulled customer data — including sales call excerpts and support conversations —...
An AI Agent Was Left Alone for Five Days — It Came Back With 190,000 Stolen Gaming Logins
Criminals are pointing a leaked AI pentesting tool at vulnerable servers and barely touching the keyboard afterward. In two campaigns documented by Check Point, autonomous agents stole over...
A Hacker’s Exposed Server Reveals How AI Chatbots Were Weaponized to Breach Korean Banks
An open directory left behind on a hacking campaign's own Hong Kong jump server exposed the AI tooling — including a hijacked Claude Code setup rerouted to DeepSeek...
DreamHost Confirms Breach After Attackers Exploited a File Manager Flaw
Hosting provider DreamHost has confirmed that attackers broke into its browser-based File Manager tool and made off with FTP/SFTP logins and some webmail credentials, with a handful of...
WorkOS Says an AI-Automated Hacking Tool Breached Its Sales Systems and Exposed Customer Slack Chats
Identity infrastructure provider WorkOS has disclosed that attackers used an AI-powered offensive security framework known as ARTEX to break into an internal sales application, exposing customer contact details...
Chainalysis Breach Exposes Customer Data in Attack Using AI-Driven ‘Agentic’ Techniques
Blockchain intelligence firm Chainalysis has confirmed an attacker used 'offensive agentic techniques' to breach internal systems, exposing customer names, emails, and role labels. The company says investigation content...