China is preparing new rules that would restrict foreign access to its most advanced AI models at the same time that America is moving to address foreign access to controlled American technology through cloud computing.
According to a Reuters report, sources say Beijing’s plans specifically target China’s top-tier, frontier-level AI systems, rather than the broader lineup of Chinese-developed AI models. The move comes as Chinese AI firms have expanded their global footprint over the past year.
Companies including DeepSeek, Alibaba, and Zhipu have released open-weight models — meaning the underlying code can be downloaded, run, and modified by anyone — that have seen widespread adoption internationally.
Chinese open-weight models overtook U.S. models in total downloads on the platform Hugging Face by August 2025, and accounted for seven of the top ten most-downloaded models on the platform between November and December 2025. Alibaba’s Qwen model family alone surpassed 1 billion cumulative downloads on Hugging Face by March 2026.
Breitbart News reported earlier this month that China’s “Kimi K3” AI model largely matched top U.S. AI models by many performance metrics:
According to testing by AI evaluator Arena, Kimi achieved superior results in front-end coding tests compared to Anthropic’s Fable 5 and OpenAI’s GPT-5.6 Sol. In broader text ranking assessments, the Chinese model performed better than Anthropic’s Opus 4.8, which served as that company’s flagship offering until Fable 5 launched in June. Notably, Kimi accomplishes these results while costing 40 percent less than comparable American models.
Moonshot AI has announced plans to release Kimi as an open-weight model on July 27. This approach differs from many premium American models by allowing companies and governments to customize the system and operate it on their own infrastructure rather than relying on cloud-based services controlled by the developer.
Chinese state media and officials have not yet detailed the specific scope of the proposed overseas access restrictions.
In the United States, separate legislation addressing a related issue — foreign access to controlled American technology through cloud computing — has been moving through Congress.
The Remote Access Security Act (RASA), authored by Sens. David McCormick (R-PA), Ron Wyden (D-OR), Tom Cotton (R-AR), and Chris Coons (D-DE) is awaiting consideration from a Senate committee. The bill would give federal regulators new authority to license or restrict remote, cloud-based access to export-controlled U.S. technology.
The bill’s scope has drawn scrutiny from some legal and technology scholars. Clark D. Asay, the Terry Crapo Professor of Law at Brigham Young University, has said that while “proponents have legitimate reasons to worry about adversaries accessing American computing infrastructure remotely,” the bill “as currently written” is “broad enough to burden allies and commercial users who pose no meaningful risks, leaving them little choice but to build their own solutions or turn to non-American AI providers.”
Oren Etzioni, professor emeritus of computer science at the University of Washington and founding CEO of the Allen Institute for Artificial Intelligence, has raised a similar point, writing that “if RASA passes, customers in other nations who currently patronize the U.S. may instead begin patronizing China,” costing the U.S. both economic activity and, in his words, “technological and political influence in what is likely the most important contest of the 21st century.”
“Barring legitimate national security concerns, the U.S. should allow American tech and AI companies to compete in the global market,” he continued. “Export controls will always have a place. But they cannot become America’s entire AI strategy.”
The Trump administration has generally embraced that broader principle, using the strength and scale of American technology companies to compete with China rather than relying on restrictions alone. Its approval of Hewlett Packard Enterprise’s acquisition of Juniper Networks through a settlement, for example, was reportedly completed to strengthen an American competitor to Huawei in the global networking-equipment market.
“In light of significant national security concerns, a settlement … serves the interests of the United States by strengthening domestic capabilities and is critical to countering Huawei and China,” a senior national security official reported to Axios.
The administration has taken a similar approach to AI. Under an executive order signed in July 2025, the Commerce Department launched the American AI Exports Program, inviting U.S. companies to submit proposals for full-stack AI technology packages to be deployed with international partners. Officials promoted the effort at the India AI Impact Summit in February, where White House Office of Science and Technology Policy Director Michael Kratsios said American AI is “the gold standard” and that the U.S. is “sharing it with our partners to secure our shared future.”
The bill has not yet been scheduled for a Senate floor vote. Its fate could shape how far U.S. export policy extends into cloud computing at the same time Beijing is weighing its own new limits on who, outside China, gets access to the country’s most advanced AI systems.
The battle for control of AI between America and China continues to heat up. Breitbart News social media director Wynton Hall has written his instant bestseller Code Red: The Left, the Right, China, and the Race to Control AI to serve as the definitive guide on how the MAGA movement can create positions on AI that benefit humanity without handing control of our nation to the leftists of Silicon Valley or allowing the Chinese to take over the world.
Lucas Nolan is a reporter for Breitbart News covering issues of AI, free speech, and online censorship.