AI-assisted breach hit nine Mexican government agencies, report says
A single threat actor used Anthropic’s Claude Code and OpenAI’s GPT-4.1 during a months-long intrusion that hit nine Mexican government organizations, according to a new…
A single threat actor used Anthropic’s Claude Code and OpenAI’s GPT-4.1 during a months-long intrusion that hit nine Mexican government organizations, according to a new…
Ransomware groups are using a wider range of tools to shut down security products before they launch encryption, and ESET says the trend now goes…
Google has started rolling out Device Bound Session Credentials, or DBSC, as a new Chrome security feature designed to cut down session hijacking. In plain…
Google has started rolling out Gmail end-to-end encryption support on Android and iOS for organizations that use Gmail client-side encryption. The update lets eligible employees…
WhatsApp is testing a username system that could let people start chats without exposing their phone number to new contacts. Right now, the feature appears…
Amazon Web Services has published an important security bulletin for Research and Engineering Studio, or RES, covering three vulnerabilities that could let an authenticated attacker…
]A newly disclosed jailbreak technique called sockpuppeting can push some major AI models past their safety guardrails by abusing assistant prefills. Trend Micro says the…
A new Magecart campaign has compromised 99 Magento stores by hiding the skimmer inside a tiny inline SVG element on checkout pages. Sansec says the…
A high-severity vulnerability in React Server Components can let remote attackers trigger denial-of-service conditions by sending specially crafted HTTP requests to Server Function endpoints. The…
DesckVB RAT is a remote access trojan that uses layered scripting, fileless loading, and process injection to avoid many traditional defenses. Point Wild’s LAT61 team…