ByteDance AI unit Seed forms dedicated safety and alignment department

ByteDance’s artificial intelligence research unit, Seed, has established a dedicated Model-Alignment & Preparedness department to focus on AI safety, alignment, and risk management. The initiative marks what is reportedly the first time a leading Chinese AI laboratory has created a standalone department specifically for these functions, elevating safety to the same level as model capability research.

Led by Yan Lin, who reports to Seed head Wu Yonghui, the department is launching a global recruitment drive to hire an initial group of ten researchers. Positions are based in Beijing, Shanghai, and San Jose, California. The research agenda includes developing model behavior specifications, conducting safety evaluations, implementing automated red-teaming, and addressing CBRN (chemical, biological, radiological, and nuclear) preparedness.

This move comes amid increasing international scrutiny regarding the safety of autonomous AI agents and ongoing discussions between the United States and China concerning AI governance frameworks and security risks.