As the former State Department envoy who co-led the first U.S.-China A.I. dialogue, the author fears upcoming safety talks will fail to prevent catastrophe. Despite presidential mandates to focus on A.I. risks, Chinese officials steered the May 2024 meeting away from dangers, refusing to explain domestic regulations and instead chiding the U.S. for safety concerns. China is portrayed as prioritizing strategic advantage over cooperation, likely using talks to soften chip controls while continuing to exploit U.S. models and generate propaganda. Even if Beijing takes risks seriously, domestic Chinese companies rarely disclose safety issues, and their models lag behind U.S. counterparts in critical safety performance. The author warns that China will feign seriousness to gain concessions, exploiting sincere demands from A.I. scientists and politicians. Without comprehensive safety agreements and domestic policy solutions first, diplomacy risks weakening the U.S. bipartisan consensus on leading A.I. development, allowing China to forge ahead in the global technology competition.

The coming meetings are unlikely to produce durable agreements while domestic policy remains under construction. Technical and oversight solutions must precede international translation. China may dangle safety cooperation for concessions, but its true aim remains gaining a competitive edge. The result could spark internal U.S. debates that undermine the core consensus that the United States must lead in all facets of A.I., making it easier for Beijing to dominate the global technology race.

Key figures mentioned include former State Department envoy, Chinese counterparts at the May 2024 meeting, and references to President Xi Jinping's recent comments on A.I. risk and loss of control.