Can the US and China work together on AI risks?

The United States and China are competing fiercely in the development of advanced artificial intelligence (AI) models, as the technology's potential risks and implications become more prominent. While the U.S. currently leads in AI capabilities, model research, and citations, China has been rapidly closing the gap due to restrictions imposed by the U.S. on China's access to cutting-edge technologies like advanced chips and manufacturing for these devices.

Experts, such as Samm Sacks from the Johns Hopkins School of Advanced International Studies, agree that the gap between US and Chinese models is narrowing and becoming increasingly fragile. This matter takes on greater significance ahead of Chinese President Xi Jinping's upcoming visit to the United States to meet with President Donald Trump, and as experts predict the future of AI will largely depend on the actions of both nations.

Several Chinese AI models have gained widespread attention and sparked concerns in both Washington and Beijing. Moonshot AI, based in Beijing, made headlines with the release of its Kimi K3 model, which ranked third globally in model intelligence in July. However, its ranking has since dropped to ninth, indicating a fluid AI landscape.

DeepSeek, a notable Chinese AI startup, challenged the US lead with its cost-effective R1 model and subsequent V4 model, which showcased significant improvements in knowledge, reasoning, and "agentic" capabilities.

Z.ai, another Chinese startup, has also garnered attention with its GLM-5.2 and GLM-5.3 models, which showcased impressive coding and agentic capabilities. Alibaba, a Chinese tech giant, previewed its Qwen3.8-Max model, claiming it to be the most powerful in its series and challenging Anthropic's Claude Fable and OpenAI's GPT models.

China's government has recognized the potential risks associated with AI technology, and has warned that it could pose threats to the country's political, institutional, and ideological security. Chinese leaders have emphasized the importance of accelerating AI development while also maintaining control of risks and establishing a global framework for AI governance.

In the United States, President Trump and his administration have emphasized the importance of maintaining the country's AI lead over China. The U.S. has accused Chinese AI developers of engaging in "industrial-scale distillation," a technique used to copy or imitate the capabilities of more powerful models. The FBI, National Security Agency, and Cybersecurity and Infrastructure Security Agency have accused Chinese AI developers of extracting capabilities from advanced US models through unauthorized access methods.

The dispute over AI distillation remains contentious, with Beijing rejecting the accusations and accusing the U.S. of pursuing a "monopoly of the AI industry." Nonetheless, the strategic implications of these allegations are significant, as Chinese developers can learn from US frontier models and release highly capable open-weight models that spread globally.