Washington’s AI Showdown Moves From Rivalry Toward Risk Control

Donald Trump and Xi Jinping will meet next week in Washington DC, and artificial intelligence will sit near the center of the conversation. The two leaders are competing for AI dominance, yet both governments also face the same question: how can they control serious risks without slowing the technology that may shape national power?

That tension gives the meeting unusual weight. Trump and Xi have both stressed that a slowdown in AI development is unnecessary, and the US and China see each other as the strongest argument against putting the brakes on progress. At the same time, both countries are discussing ways to report emergencies, exchange information, and avoid a dangerous surprise.

AI Competition Meets a New Safety Conversation

Trump has repeatedly said, “whoever wins AI, wins,” and he has also declared, “We’re leading China in AI.” His administration has announced an “AI Force” and plans to appoint a special envoy for AI. Trump has said he will not slow down AI industry growth, but he will look for bad actors through existing justice systems.

He has dismissed AI’s existential threats as a “hoax,” writing on Truth Social: “AI taking over the World, destroying Humanity, and all other things bad, is a HOAX.” Trump also said he would announce the AI “Czar” — “Only High I.Q. individuals need apply!”

That position sets a clear direction for the US: keep development moving, target misuse, and build government capacity around AI issues rather than impose a broad slowdown. The announcement of the “AI Force” could also be an attempt to undermine California’s plans to regulate AI companies under Governor Gavin Newsom.

China is building its own framework. The Chinese government has maintained a development plan for a new generation of AI since 2017 and aims to achieve global AI dominance by 2030. The Cyberspace Administration of China, or CAC, issues licenses and enforces AI regulations, while Chinese AI models must provide answers loyal to the regime on politically sensitive questions.

China released the third iteration of its AI safety governance framework last week, adding another piece to the policy structure that will shape its fast-moving AI sector. The framework arrives as Chinese models led by Moonshot AI and DeepSeek show they can be as powerful as models from US rivals such as Anthropic and OpenAI, often at a lower cost.

A Proposed Emergency Line Between Washington and Beijing

The most concrete area for cooperation involves communication during an AI-related crisis. The US and China are discussing the creation of a mechanism for AI dialogue, along with a new “notification mechanism” for incidents that affect national security.

US Treasury Secretary Scott Bessent met with Chinese Vice Premier He Lifeng to discuss AI and called the talks “very successful.” Bessent said, “The US has proposed that we have a notification mechanism between the two countries, and we want a shared vision of common goals and common threats.”

The US has already proposed a mechanism for exchanging information about serious risks posed by AI. The latest discussions focus on how the two countries would report and communicate if an AI-related emergency emerged, especially when cyber capabilities or loss of control created risks for national security.

Michael Kratsios, Trump’s science and technology adviser, said, “The key thing that we want to talk to them about is avoiding shared risks, and that’s something that I think we can have a fruitful conversation on.” Alison Szalwinski said, “We’re at a point where both the United States and China appear to agree on some of the risks posed by AI, particularly around issues of cyber capabilities and loss of control.”

China’s foreign ministry has warned that “Fearmongering, confrontation and vicious competition will only disrupt the process of global AI governance.” Xiao Qian said, “We cannot expect meaningful cooperation on AI safety if one side increasingly treats the other side’s technological progress as a national security threat.”

Trust Remains the Hardest Technical Problem

Cooperation will not erase the rivalry. The US and China are also discussing trade restrictions on high-performance AI chips and rare earth elements, two areas tied to the hardware and supply chains that support advanced systems.

The US claims China is ripping off American AI technology, a charge Chinese officials have rejected. Chinese officials and some analysts argue that “distillation” of AI capabilities is not uncommon and is not the main reason Chinese models have reached their current level.

Anthropic accused Chinese AI models of routing user requests to its Claude AI models. At the same time, Chinese AI models are making inroads globally, including in the US, creating another layer of competition between companies and governments.

Those disputes make the proposed notification mechanism difficult to build. A system for sharing information about emergencies requires both sides to trust the information, protect sensitive details, and accept that the other country’s AI progress cannot automatically be treated as a national security threat.

The Washington meeting will test whether that narrow trust can survive the wider contest. Trump and Xi do not need to settle the future of AI in one conversation, but they will need to decide whether competition and safety can occupy the same table. If they can, the next phase of AI may bring faster systems alongside clearer channels for responding when those systems create danger.

Stay connected via Google News

Follow us for the latest AI News