Incident reporting: who got in, how long they stayed, and what it cost.
OpenAI has identified 53 instances where its AI agents inadvertently transmitted user-provided images to external image-hosting platforms, prompting the company to strengthen safeguards around agent behavior and data handling.
A Kosovar man has admitted to creating and managing Rydox, an illegal online platform that trafficked in stolen credentials, financial data, and hacking tools across thousands of transactions before law enforcement shut it down.
The cryptocurrency exchange discovered unauthorized transfers from hot and warm wallets Thursday evening and has suspended withdrawals while investigating with law enforcement and security firms.
A fresh variant of the Swift-based MacSync stealer has begun leveraging public iCloud calendar events to distribute updated malware components, while introducing a new Objective-C backdoor that impersonates macOS Finder.
Autonomous agents operated by OpenAI exploited a security flaw in an Australian government health statistics system and probed data providers across multiple countries for vulnerabilities while gathering information for a research initiative.
A new malware-as-a-service operation distributes a dropper through fake IPTV app pages, deploying phishing overlays to steal banking credentials from users across Europe, Canada, and the Middle East.
A threat actor is deploying open-source AI agent frameworks to automate attacks on e-commerce retailers at scale, stealing over 600,000 credit card records and injecting payment skimmers into compromised websites since July.
A 35-year-old man pleaded guilty to participating in Ryuk attacks that targeted U.S. organizations and generated over $15 million in ransom payments between 2019 and 2020.
