Seed Laboratory, a subsidiary of ByteDance, has recently established a new department called the "Seed Model-alignment & Preparedness" division. This new department is headed by Yan Lin and reports directly to Wu Yonghui, the leader of Seed. It brings together members from the former Responsible AI team and segments of the Chat team, with the goal of enhancing research and investment in model safety and alignment. The new department will prioritize improving the safety features of intelligent agent models, safeguarding against potential risks associated with cutting-edge capabilities, including those in the chemical, biological, radiological, and nuclear (CBRN) sectors. Its responsibilities encompass developing model specifications, constructing safety assessment systems, conducting alignment training, establishing protection mechanisms, and monitoring risks. Additionally, it will foster collaboration with external research institutions and third-party evaluation organizations. As the first cutting-edge AI laboratory in China to create a dedicated department for safety and alignment, Seed's restructuring aims to establish safety and alignment as distinct functions. This ensures that cutting-edge models can effectively carry out tasks within safety regulations and authorization limits, mitigating the risks of model misuse and behavioral deviations, and enhancing the safety framework to tackle unknown risks.