The joint action of building large model security evaluation standards has officially started.

The National Industrial Information Security Development Research Center is building a large-scale model security assessment benchmark based on the technical accumulation and practical experience in the field of large-scale model security assessment. The center is now soliciting co-construction units for the large-scale model security assessment benchmark, aiming to create a comprehensive and systematic artificial intelligence security assessment system. The large-scale model security assessment benchmark being developed this time is focused on the security governance requirements in the real application of large-scale models, constructing a panoramic evaluation framework covering six core security dimensions: content security, value alignment, robustness, fairness and impartiality, privacy protection, and authenticity and trustworthiness, forming a systematic end-to-end evaluation system.