New Act Demands Detailed Annual Report on China’s AI Power, Focusing on Chips and Military Models

This new legislation, the China AI Power Report Act, isn’t about changing what you can buy or sell today. It’s about intelligence gathering—specifically, requiring the U.S. government to get a serious, annual check-up on exactly where China stands in the race for artificial intelligence dominance. Think of it as demanding a comprehensive, 15-point diagnostic report on a competitor’s engine.

The bill mandates that the Secretary of Commerce produce a detailed annual report for the next three years, consulting with agencies like the Departments of State and Defense, and the Director of National Intelligence. The core purpose, as stated in the bill (Sec. 2), is to give Congress the hard data it needs to keep U.S. export controls "dynamic and adaptive" against China’s evolving capabilities. In short: they want to plug the loopholes before they become problems.

The Semiconductor Deep Dive

If you’ve heard anything about the tech war, you know it’s all about the chips. This report zeroes in on the hardware that powers modern AI. The Commerce Secretary must assess everything from advanced AI chips designed by Chinese firms like Huawei, down to the logic and memory chip fabrication facilities (like SMIC and CXMT) that make them. They need to know production numbers, yields, what equipment is being used, and where that equipment originally came from (Sec. 3).

This isn't just academic. If you work in a sector that relies on global supply chains—from automotive to consumer electronics—understanding who controls the production of advanced chips is critical. This report aims to map out the entire semiconductor ecosystem, including chip manufacturing equipment and the electronic design automation (EDA) software used to design the chips in the first place.

Mapping the AI Brainpower

The report also demands a comprehensive look at the actual AI being developed. This includes assessing the most relevant AI models coming out of Chinese labs, especially those linked to the People’s Liberation Army (PLA). They want to know the models’ capabilities, the compute infrastructure they run on, and how they are being used across Chinese society and the military.

Crucially, the report must also track the money, requiring an assessment of total public funding and capital flows supporting AI development in China. For everyday people, this means policymakers will have a clearer picture of whether U.S. capital or investments are inadvertently fueling technologies that could pose a national security risk down the line. It’s about ensuring U.S. dollars aren't underwriting foreign strategic advantages.

Tracking the Loopholes and Leakage

Perhaps the most actionable part of the report is the focus on control effectiveness. The bill requires an assessment of how Chinese entities remotely access AI computational resources (like cloud services) to get around U.S. controls. Furthermore, it demands an analysis of the methods and quantities of U.S.-controlled AI chips diverted to China and what percentage of China’s total compute capacity that represents. This is the government admitting they know diversion is happening and they need the numbers to figure out how to stop it (Sec. 3).

For those of us working in tech or manufacturing, this means future export controls are likely to become more targeted and sophisticated, based on this detailed intelligence. The report’s unclassified portion must publicly include the number of advanced AI chips produced in China and the projected numbers for the coming year, giving the public a rare, hard look at the scale of this technological competition.