OpenAI, Anthropic, Google, and xAI have united in warning that artificial intelligence could slip beyond human control, and the U.S. government is now trying to make AI safety a centerpiece of high-level diplomacy with China. As President Donald Trump prepares to meet Chinese President Xi Jinping at the White House on September 24, U.S. Treasury Secretary Scott Bessent has signaled that Washington wants to discuss AI safety cooperation with Beijing. However, experts and diplomatic observers say the chances of China accepting such an offer are slim.

The upcoming summit represents a rare moment when the world's two largest AI powers could theoretically align on safety standards. Leading U.S. AI developers, including OpenAI (maker of ChatGPT), Anthropic (builder of Claude), Google (with Gemini), and xAI (creator of Grok), have recently spoken in unison about the dangers of advanced AI systems. Bessent told U.S. media outlets on September 16 that Washington is open to talks aimed at avoiding risks both sides face in common and at preventing the two countries' AI systems from decoupling from each other.

Why Is AI Safety Suddenly a Diplomatic Priority?

The shift toward making AI safety a diplomatic issue reflects growing concern among leading AI companies about the trajectory of their own technology. Developers have argued for what is known as "slowing the pace," warning that without proper safeguards, AI could eventually operate beyond meaningful human oversight. This concern has moved from academic circles and company research labs into the halls of government, where officials now see it as a geopolitical issue that requires coordination between the world's AI superpowers.

The timing is significant. Unlike previous Trump-Xi meetings in October 2025 and May 2026, which focused on rare earths, soybeans, semiconductors, the Middle East, and Taiwan, this summit is expected to elevate AI safety to a central topic. The U.S. Treasury Secretary, Trade Representative Jamieson Greer, and Chinese Vice Premier He Lifeng are scheduled to meet in New York on September 20 to coordinate the summit agenda, suggesting that AI safety discussions are being taken seriously at the highest levels.

What's Beijing's Position on AI Safety Cooperation?

Despite U.S. efforts to frame AI safety as a shared concern, China views the push differently. Beijing interprets AI safety regulation and calls to slow AI development as a U.S. strategy designed to constrain China's own AI industry and maintain American technological dominance. This fundamental disagreement makes cooperation unlikely, according to diplomatic observers.

China's response has been to build alternative alliances and pursue an open-source approach to AI development. In July 2026, China joined Russia and other nations in establishing the World Artificial Intelligence Cooperation Organization, explicitly created as a counter to the U.S.-led Pax Silica initiative. Pax Silica is an AI supply chain body created in December 2025 under U.S. State Department leadership to restrict China's access to advanced AI chips and technology. South Korea is one of its main participants.

China's major AI models, including Moonshot AI, DeepSeek, Alibaba, MiniMax, and Z.AI, are also expanding their reach through open-source approaches, unlike their U.S. counterparts, which are typically closed systems available only through commercial licensing. This strategic difference means that even if the two countries agreed on safety standards, enforcing them would be far more difficult in China's decentralized AI ecosystem.

How Might the Summit Actually Play Out?

Diplomatic circles suggest that a successful outcome would be modest in scope. Rather than achieving a comprehensive AI safety agreement, experts expect the two countries might agree on minimal institutional safeguards designed to prevent a head-on clash over AI regulation. Trump himself has indicated some ambivalence about the safety-first approach. Speaking to reporters in Dublin on September 13, Trump said he wants to stay ahead of China in AI and that while safeguards can be put in place, there are "plenty of negative forces raising things that need not be raised".

Trump

Beyond AI safety, the summit will address several other contentious issues. The U.S.-China trade truce, which was built around Chinese rare earth exports and soybean purchases in exchange for U.S. tariff exemptions and permission for some AI chip exports, is set to expire on November 10. Both sides are expected to extend it, though they disagree on the length. Washington wants a six-month extension, while Beijing is pushing to set the term at the remainder of Trump's presidency.

The Trump administration has also delayed so-called overproduction tariffs on major trading partners, including China, until after the summit. The administration had initially set the overproduction tariff rate on Chinese goods at 7.5%, according to Bloomberg reporting on September 17. China has already been hit with a 12.5% forced labor tariff on top of existing steep duties, imposed on July 24 under Section 301 of the Trade Act.

Steps to Understanding the AI Safety Debate in U.S.-China Relations

• Know the Players: OpenAI, Anthropic, Google, and xAI are the primary U.S. AI developers pushing for safety standards, while China's Moonshot AI, DeepSeek, Alibaba, MiniMax, and Z.AI represent the competing Chinese approach to AI development and deployment.
• Understand the Philosophical Divide: The U.S. emphasizes closed, carefully controlled AI systems with built-in safety measures, while China favors open-source models that are more widely distributed and harder to regulate centrally.
• Recognize the Geopolitical Context: AI safety discussions are inseparable from broader U.S.-China competition over technology dominance, trade, and supply chain control, making pure technical cooperation difficult.
• Track the Institutional Landscape: The U.S.-led Pax Silica initiative and China's World Artificial Intelligence Cooperation Organization represent competing frameworks for how AI development should be governed globally.

The September 24 summit will reveal whether Trump administration officials can convince Beijing that AI safety is a shared concern rather than a disguised effort to slow Chinese technological progress. Many observers remain skeptical. He Yadong, a spokesperson for China's Ministry of Commerce, said on September 17 that the two sides are continuing talks on tariff cuts worth $30 billion, suggesting that trade issues will likely dominate the agenda over AI safety discussions.

What happens at the White House on September 24 could set the tone for how the world's two largest AI powers approach safety, regulation, and competition in the years ahead. If the summit produces only a vague commitment to continued dialogue on AI safety, it may signal that the fundamental disagreement between Washington and Beijing on how to govern AI development remains too deep to bridge through diplomacy alone.

" }