Daily curates a variety of tech resources, tools, tips, and news (5G, big data, cloud computing, AI), aiming to become a go-to popular science encyclopedia for everyone.
This article explains how to move threat intelligence from static IOC collection into a five-step closed loop—collect, normalize, correlate, remediate, review—integrating with assets, vulnerabilities, logs, and ticketing to prioritize real risks, avoid alert fatigue, and enable actionable security operations.
This article outlines six essential security boundaries enterprises must verify before deploying AI Agents, including scenario, data, tool, action, audit, and operational limits, plus a risk matrix and layered defense framework to prevent tool-calling capabilities from becoming uncontrolled business risks.
China's new mandatory cybersecurity labeling system for consumer network cameras introduces a three-tier rating that shifts IoT security from reactive patching to verifiable, comparable trustworthiness, giving buyers a five-look procurement model and manufacturers six required capabilities while users get a seven-step self-audit checklist.
China's new Network Data Security Risk Assessment Measures, effective August 2026, shift focus from static compliance to dynamic risk management, requiring organizations to identify data assets, map flows, define concrete risk scenarios, and provide remediation evidence across six key scenarios including AI training and third-party processing.
The article argues that successful AI deployment depends on robust data governance — usable, trustworthy, and circulatable data — rather than model upgrades, citing China's 'Data Elements ×' policy progress and the rise of autonomous agents that demand traceable, permissioned data foundations.
This article argues that successful AI agent deployment depends not on model size but on coordinated digital infrastructure—network, compute, edge, data, and security—citing China's 2026-2028 policy, agent capabilities (context, tools, integration, auditability), dynamic compute placement, data governance, and three maturity criteria: stability, process integration, and compliance.
As AI enables low-cost, hyper-realistic fraud across familiar digital channels, traditional awareness campaigns fall short; this analysis argues for a four-layer verification framework—source, identity, transaction, and platform—to replace passive alerts with actionable defenses, backed by China's 2026 regulatory crackdown on unlabeled AI content.
China's MIIT releases a 2026-2028 policy to embed AI into communication networks, shifting from application-layer AI to intelligent infrastructure covering network autonomy, edge computing, fraud prevention, and integrated data-network-compute-scenario loops for real-time industrial applications.
This article details six network flow analysis methods—timeline reconstruction, DNS/IP/TLS profiling, app behavior reconstruction, multi-case correlation, and scoring-based warning models—to help investigators trace fraud infrastructure beyond traditional person-card-money leads, emphasizing evidence compliance and analytical rigor.
The article maps 2026 AI tools into seven practical categories — general assistants, deep research, office-integrated AI, coding agents, visual generators, meeting transcription, and automation agents — and argues selection should prioritize data integration, task reliability, error traceability, and workflow embedding rather than chasing benchmark leaders.
