Category: AI Security

Nogosee's tracker-first policy allows monitor-only records to remain valuable through scoring, filtering, and brief queues without forcing weak signals into full articles. This workflow guides security, cloud, and operations teams on how to use the East Asia Cyber & AI Risk Tracker for repeatable monitoring, verification, and escalation while preserving source integrity and avoiding low-value publishing. Read more

This guide provides a step-by-step workflow for security teams to assess whether a critical-infrastructure signal from Nogosee’s East Asia tracker warrants inclusion in a regional risk brief, focusing on source scope, sector relevance, operator type, dependency chains, and uncertainty boundaries without relying on numeric thresholds or rigid rules. Read more

MCPThreatHive is an open-source platform that automates MCP threat intelligence through continuous data collection, AI-driven extraction, and a unified taxonomy of 38 threat patterns mapped to STRIDE and OWASP frameworks, addressing gaps in compositional attack modeling and continuous monitoring. Read more

This checklist guides analysts in extracting actionable intelligence from public cyber incident disclosures using Nogosee’s East Asia Cyber & AI Risk Tracker. It outlines steps for identifying source wording, affected entities, sectors, uncertainty levels, response status, and watchlist follow-up, with clear ownership, decision criteria, and escalation paths for security and operations teams. Read more

CVE-2026-45829 in ChromaDB’s Python FastAPI server allows unauthenticated remote code execution by executing malicious models from Hugging Face before authentication verification, affecting an estimated 73% of exposed instances and posing significant risk to agentic AI deployments reliant on dynamic model loading. Read more

GovCERT.HK's High Threat Security Alert (A26-05-36) confirms active exploitation of CVE-2026-42945, a denial-of-service and remote code execution vulnerability in Nginx, affecting stable versions prior to 1.30.1 and mainline prior to 1.31.0. The alert references six CVEs spanning memory safety, input validation, and configuration flaws, with potential impacts including full system compromise, service disruption, and data exfiltration. Given Nginx's pervasive role as a reverse proxy, load balancer, and ingress controller in enterprise, cloud, and containerized environments, the vulnerability presents a critical initial access vector for threat actors. Immediate patching is urged, with defenders advised to verify patch status across public-facing, internal, and cloud-deployed instances while monitoring for exploit indicators in logs and network traffic. Read more