aiqbench analysis
NetEase Smart Enterprise has launched a commercial security suite aimed at risks in deployed AI systems, covering content and data protection, identity and interaction security, and model and agent security. The announced tools include a review agent and face-forgery detection, but the article gives no performance data, benchmarks, or independent validation, and the "Debut Achievement Certificate" is a showcase award rather than a technical assessment. Readers should treat the claims as vendor positioning until there are measurable results.
NetEase Smart Enterprise Unveils AI-Native Security Product System Built on Dun Shield's Two-Way Architecture
36Kr learned that on November 3, NetEase Smart Enterprise unveiled an AI-native security product system based on the two-way architecture of Dun Shield at the 2026 World Internet Conference. The system has received a "Debut Achievement Certificate" issued by the Internet of Light Launch Hall. The system covers three layers: AI content and data security, AI interaction and identity security, and AI model and agent security. Three achievements were also released: the CMA review agent, DeepGuard face forgery detection, and AI application and agent security.
Read the full article at 36KrRelated coverage
A Reference Architecture for Integrating Agentic AI into Enterprise Software: Responsibilities, Interactions, and Variability
arXiv · Oct 15, 2026
.webp%3Fssl%3D1&w=320)
OpenMatter Network Calls on Enterprise Leaders to Rethink AI Security Before the Next Rogue AI Crisis
Cybersecurity News · Jul 30, 2026
Scaling agentic AI: Enterprise patterns without vendor lock-in
AWS · Aug 20, 2026
What’s new with Google Cloud
Google Cloud Blog · Oct 30, 2026