The United States and China are set to launch an official AI safety dialogue to establish "guardrails" in their technological competition. U.S. Treasury Secretary Scott Bessent will lead the American delegation, while Chinese Vice Minister of Finance Liao Min has been involved in preliminary talks. The dialogue aims to create a crisis management mechanism akin to Cold War-era strategies, focusing on preventing unpredictable AI behavior, controlling autonomous military systems, and addressing threats from non-state actors.Despite the initiative, challenges loom due to Bessent's hardline stance on China-related AI issues, which may complicate negotiations. Historically, previous attempts at AI dialogue have been limited, but non-governmental communication channels remain active, fostering ongoing discussions on AI safety and alignment.