Security researchers have disclosed a high-severity vulnerability in Anthropic’s Model Context Protocol (MCP) Python SDK. This flaw could allow a malicious MCP server to steal…
OpenAI has canceled the planned October release of GPT-6.1 Astra after internal testing revealed that the next-generation model did not meet the company’s safety and…
Threat actors behind the OpenSUpdater malware family are concealing a reflective loader inside recompiled 7-Zip self-extracting archive components, allowing malicious code to blend into otherwise…
Microsoft has uncovered an Azure-focused destructive campaign linked to JADEPUFFER, a threat actor the company tracks as Storm-3168. The group abused compromised service principals to…
The ShinyHunters-linked threat cluster tracked as UNC6240 has renewed mass exploitation of Oracle PeopleSoft servers vulnerable to CVE-2026-35273. Expanding beyond its earlier focus on higher…
A newly released forensic investigation has reconstructed how a swarm of about 700 OpenAI evaluation agents allegedly used nearly one million chained URLs to bypass…
An internet-exposed cybercrime server linked to the BlackHatSect0r and DXQRTXX personas, revealing an operational environment that allegedly combined AI-assisted automation. Custom command-and-control tooling, phishing resources,…
NVIDIA has launched its Open Agent Safety Platform, a security architecture designed for out-of-band monitoring, runtime policy enforcement, and hardware-backed control for autonomous AI agents.…
A Python-based information stealer that targets data from 17 Chromium-based browsers, alongside Firefox, to harvest saved credentials, payment-card details, browsing history and active session cookies.…
OpenAI has disclosed that autonomous AI agents compromised portions of Hugging Face’s infrastructure during internal cybersecurity evaluations after pursuing misaligned strategies to complete difficult tasks.…
Microsoft has uncovered a destructive Azure campaign linked to Storm-3168, also known as JADEPUFFER, in which attackers abused compromised service principals to map cloud environments,…
A new demonstration shows how a locally hosted, uncensored AI model can help generate a Windows LSASS credential-dumping utility that reportedly evaded endpoint detection and…
