China has achieved a significant milestone in next-generation power technology by commercializing the world's first supercritical CO2 power generator, potentially giving it a strategic advantage in powering energy-intensive AI systems and military operations. On December 20, 2025, the China National Nuclear Corporation (CNNC) announced that Chaotan One, a joint project with the Nuclear Power Institute of China (NPIC), began commercial operations as a 2×15 megawatt (MW) supercritical carbon dioxide waste-heat power generation demonstration project.

What Makes Supercritical CO2 Technology Different?

Supercritical CO2 (sCO2) is created when carbon dioxide gas is pressurized and heated to conditions where it possesses the density of a liquid while maintaining the expansive properties of a gas. This unique state allows sCO2 to generate electricity more efficiently than traditional steam-based systems. In a closed-loop sCO2 Brayton cycle, compressed sCO2 is heated, expanded through a turbine to generate energy, and then recompressed.

The practical advantage is striking: sCO2 turbines can be a fraction of the size of traditional steam systems while delivering comparable or superior power output. This directly translates to nuclear and fossil fuel plants operating more efficiently at lower costs with a smaller physical footprint, making the technology particularly valuable for applications where space and resource constraints are critical.

Why This Matters for AI and Military Power Demands?

As artificial intelligence systems consume ever-increasing amounts of electricity, compact and efficient power generation has become a national security priority. The U.S. military is increasingly deploying energy-intensive assets, including directed-energy weapons deployed in a federal pilot in May 2026 and AI systems for target recognition announced during the same period. In austere or contested environments, traditional power systems face significant vulnerabilities. Diesel fuel and water supplies are easily targeted for interdiction, but sCO2 systems reduce fuel demand and require significantly less cooling water than traditional steam cycles, mitigating two of the military's most easily targeted dependencies.

A modular sCO2 turbine could provide surge capacity for energy-intensive military and AI systems without requiring an expanded physical or thermal footprint, making it ideal for forward-deployed operations and data center applications where space is limited.

How to Understand China's Strategic Advantage in sCO2 Development

• Research Foundation: Chinese universities and state-owned enterprises have been publishing peer-reviewed research on supercritical CO2 since at least 2017, with examples including China University of Petroleum and Sinopec-affiliated research, establishing a decade-long knowledge base.
• Testing and Validation: In early 2019, Chinese concentrated solar power manufacturer Shouhang partnered with French utility EDF to retrofit a 10 MW solar plant with an sCO2 power cycle, marking the first operational test of sCO2 technology in a commercial facility and proving its commercial viability.
• Government Priority: China's Fourth National Communication on Climate Change in late 2023 highlighted sCO2 as a critical technology item to support energy development, including uranium mining for nuclear energy use, demonstrating sustained policy support.
• First-Mover Advantage: Chaotan One's successful commercialization positions China as the leader in industrial sCO2 deployment, potentially allowing Beijing to dominate the supply chain for next-generation power components and compact machinery.

The timing of China's sCO2 breakthrough is significant given the country's existing participation in building critical energy infrastructure for emerging economies as part of the Belt and Road Initiative. If Chaotan One proves successful long-term, this first-move advantage could provide China with substantial strategic leverage in global energy markets.

How Is the U.S. Responding to the Nuclear Energy Challenge?

The United States has launched an aggressive response to accelerate its own advanced nuclear development. The Department of Energy (DoE) announced the Nuclear Pilot Program with a goal of starting operations for at least three test reactors by mid-2026. More recently, the DoE set a July 4, 2026 deadline for at least three advanced reactors to achieve criticality, a self-sustaining nuclear chain reaction that validates years of design and engineering work.

Two projects have already crossed this milestone. Antares Nuclear's Mark-0 reactor achieved criticality on June 4, 2026, becoming the first U.S. non-light-water reactor to do so in more than four decades, while Valar Atomics followed on June 18 with its Ward 250 microreactor. Aalo Atomics is expected to become the third reactor to reach criticality before the July 4 deadline.

U.S. policymakers have expressed confidence in the timeline, with officials describing the current period as the beginning of a new phase for advanced nuclear development supported by faster regulatory processes and stronger policy backing. Some small modular reactors (SMRs) could begin generating electricity as early as 2027, with wider commercial deployment expected before 2028.

The Department of Defense and Department of Energy have also collaborated on transporting a nuclear reactor via C-17 cargo aircraft, while the DoE announced up to $800 million in advanced deployment funding for advanced light-water small modular reactors in December 2025. Additional initiatives include focused efforts to develop advanced nuclear fuel lines and the creation of a Defense Production Act Consortium focused on domestic nuclear fuel cycle supply chains.

How Are AI and Nuclear Energy Becoming Integrated?

The convergence of AI and nuclear power is accelerating through government-backed initiatives. The Department of Energy's Genesis Mission unites National Labs, industry, and academia to harness AI for breakthroughs in energy, science, and national security. AWS is collaborating with Idaho National Laboratory (INL) to use AI to develop a digital twin of a small modular reactor, ultimately enabling what INL calls "nuclear energy AI at scale".

This collaboration involves using AWS technology to compress nuclear reactor design cycles, generate digital twins for advanced simulation, and advance autonomous reactor operations. The National Nuclear Security Administration (NNSA) announced a significant milestone by establishing a Secret/Restricted Data (S/RD) Enterprise Cloud environment in collaboration with AWS, marking the first cloud environment to receive enterprise authorization to process classified nuclear data and house the NNSA's inaugural Genesis Mission workloads.

"Technology is a top priority," said CIA Director John Ratcliffe in his first public remarks in office, confirming that the CIA will take advantage of cloud modernization programs to dramatically strengthen the backbone of its entire IT architecture, with acquisition timelines cut from 12-24 months to under six months and the agency going all-in on AI.

John Ratcliffe, CIA Director

The broader policy environment reflects recognition that advanced nuclear technology and AI are becoming inseparable. The U.S. administration ultimately wants nuclear capacity to reach roughly 400 gigawatts by 2050, with advanced reactors expected to serve military bases, data centers, industrial facilities, and export markets. The DoE is also expanding testing infrastructure through initiatives such as the Nuclear Energy Launch Pad to provide developers with a more permanent pathway from demonstration to commercial deployment.

For investors and policymakers, the investment case is no longer centered on a single reactor announcement but rather on a policy framework designed to shorten development timelines and encourage private investment. However, commercial deployment still requires additional testing, licensing, and execution, making regulatory progress and project delivery important factors to monitor alongside technological advances.