China has developed and launched a powerful artificial intelligence (AI) system that connects directly to its national supercomputing network. This system is capable of conducting advanced scientific research independently.
The initiative started on December 23, just a month after US President Donald Trump announced the Genesis Mission, which is an “AI Manhattan Project” intended to maintain US technological leadership.
While the US plan is under strict deadlines that require proof of progress within 270 days, China has already rolled out its system across the country, serving over a thousand potential institutional users.
The new AI agent can understand simple language commands. It can break down complex research tasks, allocate computing power, run simulations, analyze data, and generate complete scientific reports without needing constant human oversight, according to the official China Science Daily.
This system operates on China’s National Supercomputing Network (SCNet), a fast digital framework that connects more than 30 supercomputing centers throughout the nation.
It supports nearly 100 scientific workflows in areas like materials science, biotechnology, and industrial AI.
What used to take scientists a full day can now be completed in about an hour, according to SCNet.
This network may change the dynamics of the US-China tech competition and have broader implications worldwide.
For decades, both countries have spent billions in developing supercomputers that simulate nuclear explosions, design next-generation stealth fighters, and discover life-saving drugs.
A major question in the tech community has been whether AI might gain access to sensitive and often classified data, effectively turning it into a “super scientist.”
Allowing AI direct access to national supercomputing networks could significantly speed up scientific breakthroughs, but it also poses substantial risks.
These risks include potential leaks of government data about citizens, enabling hackers to gather sensitive information, or giving AI access to details about military systems.
“Science is shifting from number crunching to AI-driven discovery. These new AI agents will connect tools, data, and computing resources spread across different systems, providing scientists with better tools to innovate quickly,” said Qian Depei, a member of the Chinese Academy of Sciences and head of an SCNet expert panel, at the project’s press conference.
“AI for Science is not just a technical path but also a change in how research is structured,” stated Cao Zhennan, deputy director of the High Performance Computer Research Centre, during the event.
In August, Beijing introduced its national “AI+” initiative, explicitly promoting the use of AI to speed up scientific discovery.
The Genesis Mission was Washington’s response, with plans to leverage federal supercomputers and decades of government research data to train robust AI agents.
During the launch of the project, Trump stated it aimed to “invest in AI-enabled science to quicken scientific progress” and enhance “America’s technological leadership and global strategic influence.”
Led by the Department of Energy, the mission will create an American Science and Security Platform that will use the US government’s scientific data sets— the largest in the world—to train AI models and develop intelligent agents.
These agents will be tasked with testing new hypotheses, automating research workflows, and accelerating scientific discovery.
The platform will utilize extensive government supercomputers to train AI models and produce intelligent agents. These agents will test hypotheses, automate research tasks, and hasten breakthroughs.
The US plan has a rigid timeline. Within 60 days of starting, it must identify 20 national science and technology challenges that the mission could tackle, and it must demonstrate operational ability within 270 days.
Six priority areas have been pinpointed: advanced manufacturing, biotechnology, critical materials, nuclear fission and fusion energy, quantum information science, and semiconductors and microelectronics.
China’s SCNet was launched in 2023 and aims to merge supercomputing and intelligent computing resources nationwide. It connects supercomputing centers in a fast network that supports national development.
The platform was officially unveiled in the Binhai Hi-Tech Zone in Tianjin in April 2024 and has since linked over 30 computing centers, which cater to users from more than a thousand government agencies, companies, universities, and research institutions.
