Track East Asia cyber, AI, cloud, and infrastructure risk before it becomes an incident.
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
A new arXiv paper reveals that stringent university cybersecurity measures—such as MFA, device compliance, and remote management—disproportionately block international students in UK-China transnational programmes due to time-zone gaps and lack of real-time IT support, exposing a critical flaw in co-located security assumptions. Read more
MolTrust implements a production trust infrastructure for autonomous AI agents using W3C Verifiable Credentials and DIDs, with 69,000 bots processing 165 million transactions worth $50M USDC on Base Layer 2 since March 2026, featuring kernel-level authorization enforcement and cross-protocol interoperability. Read more
Researchers developed TWGuard, an LLM safety guardrail optimized for the Taiwan linguistic context, achieving a +0.289 F1 gain over foundation models and a 94.9% reduction in false positive rate compared to the strongest baseline, demonstrating the value of linguistic context in AI safety. Read more
A new research paper argues that current AI safety case methodologies lack rigor by over-relying on alignment community approaches instead of established safety assurance practices from aerospace, nuclear, and automotive sectors, proposing a stronger framework grounded in decades of safety-critical system validation. Read more
Use Nogosee's public tracker to assess East Asia vulnerability signals by verifying asset exposure, supplier impact, exploitability evidence, business ownership, compensating controls, and monitoring needs before escalation. Read more
Nogosee monitored 13 Taiwan listed-company cyber incident disclosures in April 2026. The cleaned month is incident-only: 13 incident records, 0 governance records, and 0 procurement records. This is a MOPS incident-disclosure snapshot for analyst workflow use, not a complete census of Taiwan cyber activity. Read more
A practical workflow for security teams to efficiently triage East Asia cyber and AI risk signals using Nogosee’s public tracker, focusing on filtering, ranking, and decision-making without hard thresholds or numeric claims. Read more
A research study identified over 1.2 million exposed Thai National Identification Numbers via search engines, with most originating from government websites, highlighting systemic data protection failures in public sector digital systems. Read more