ByteDance Seed Establishes Dedicated Large Model Safety and Alignment Department

ByteDance Seed recently issued an internal memo announcing the creation of a new 'Seed Model-Alignment&Preparedness' department, led by Yan Lin and reporting to Seed head Wu Yonghui. The original Responsible AI team and select Chat team members will transfer to this new unit. This move aims to increase investment in safety and alignment, with core objectives including ensuring model behavior aligns with human intent and safety constraints to enhance Agent Model security capabilities and mitigate risks associated with advanced technologies such as CBRN (chemical, biological, radiological, and nuclear).