US and China Prepare for First Official AI Dialogue Under Trump

Treasury Secretary Scott Bessent is slated to lead the American delegation, though the exact agenda and participants remain in flux. The discussions follow a summit between President Donald Trump and Xi Jinping in May, signaling an effort to establish guardrails as both nations race to develop frontier AI models. Officials on both sides express concern that these systems could destabilize global cyber security, enhance military capabilities, and disrupt labor markets.

Bessent recently signaled a hardline stance on intellectual property, citing evidence of U.S. large language model watermarks appearing in Chinese systems. This focus on proprietary technology complements ongoing reviews by U.S. Trade Representative Jamieson Greer, who is examining China’s state subsidies and potential forced technology transfers. While the U.S. maintains export restrictions on high-end AI chips, it has notably held back from adding firms like DeepSeek and CXMT to trade blacklists to prevent immediate escalation.

Beijing appears focused on technical cooperation, aiming to clarify how Washington intends to regulate future models and whether open-weight systems will face new barriers. Paul Triolo, a partner at DGA-Albright Stonebridge Group, suggests that these preliminary talks will likely prioritize basic terminology and common definitions rather than immediate policy breakthroughs. The meeting represents a critical, if fragile, effort to maintain communication channels while the underlying technological rivalry continues to intensify.