The National Industrial Information Security Development Research Center is
leveraging its technical experience in large-model safety testing to build a
safety-evaluation benchmark and has begun soliciting co-construction partners.
The benchmark will deliver a comprehensive, systemized AI security assessment
framework aimed at real-world governance needs, covering six core
dimensions—content safety, value alignment, robustness, fairness, privacy
protection and authenticity—and will form a full-chain evaluation system.
