ByteDance Seed recently sent an internal email announcing the establishment of a "Seed Model-Alignment&Preparedness" department, led by Yan Lin, reporting to Seed head Wu Yonghui. The original Responsible AI team and some members of the Chat team will transfer into the new department.
According to reports, the department is intended to strengthen investment in the safety and alignment direction; the core of alignment is keeping model behavior consistent with human intent and safety constraints, thereby improving the safety capabilities of Agent Models (intelligent agent models) and guarding against frontier capability risks such as CBRN (chemical, biological, radiological, and nuclear).
Large model safety risks are becoming increasingly prominent, with incidents such as network attacks triggered by agent escape occurring frequently. Previously, the safety and alignment work of most domestic large model research teams was folded into their companies' overall security and compliance systems, whereas ByteDance Seed is the first frontier AI laboratory in China to establish a separate safety and alignment department.
The Seed team internally judges that as models become more capable and application scenarios more complex, safety and alignment are no longer just risk prevention, but a core capability that determines whether models can be applied at scale.
It is understood that the Seed Safety and Alignment department will continue to invest in directions such as model specification development, safety evaluation system construction, alignment training, protective system building, and risk monitoring, and will strengthen open collaboration with external research institutions and third-party evaluations, ultimately enabling frontier models to effectively complete tasks within safety and authorization constraints, continuously reducing the risks of misuse and behavioral deviation as model capabilities and autonomy grow, and building a comprehensive safety system for facing unknown risks.
Lei Feng Wang Lei Feng Wang (public account: Lei Feng Wang)