Xi tells Trump that China and US must manage AI for good

By Splitfeed EditorialLast updated Sep 28, 2026, 16:11 UTC

On September 24, 2026, Chinese President Xi Jinping told President Donald Trump at a White House summit that China and the United States share the capability and responsibility to develop and manage artificial intelligence for good, and that both countries must keep AI under human control. The three-day meeting, which ran through September 26 and included a state dinner and a tour of the National Archives, covered AI, trade, and Taiwan. Chinese state media subsequently framed the outcome as the United States accepting shared responsibility for AI governance, a characterization that differed from how Western outlets described the same exchange.

The divergence in framing matters because it signals that both governments are competing to define the terms of any eventual AI arrangement before one is actually reached. Xi's call for partnership rather than rivalry lands at a moment when AI capabilities are advancing quickly enough that governance gaps between the two largest AI powers carry real consequences. The central dispute is whether any meaningful agreement emerged at all, and whether Trump's posture at the summit made safety outcomes more or less likely. Trump's reported refusal to negotiate formal AI guardrails with China is the specific flashpoint, since it means the summit produced warm language without binding commitments.

The pro-AI camp, represented by reactions from Dare Obasanjo and Savchenko Volodymyr, reads the summit as a reminder that the United States must accelerate its own development rather than coordinate with Beijing, with some in that camp preferring to frame the technology as superintelligence to underscore its strategic weight. Critics are sharply opposed: Senator Chris Van Hollen argued that Trump's refusal to negotiate guardrails puts everyone at risk, Senator Andy Kim said Trump was rolling over to Xi while Americans demand safety action, and The Washington Sun reported that even Trump's own allies were privately alarmed by his conciliatory tone. The middle ground, voiced by Peter Baker and the New York Times, is that after days of meetings the leaders had little to show beyond pageantry, though Aveline Quillon noted that AI safety is gaining real diplomatic weight, and several voices in this camp argued the competition should be cooperative rather than zero-sum, with one comparison drawn to lessons from the nuclear arms race.

The argument will sharpen around whether any formal AI governance talks are scheduled as a follow-on to this summit, and whether either government publishes a joint statement or framework document. The absence of such a document by the end of 2026 would largely confirm the middle-ground reading that the summit was symbolic.