Breaking: OpenAI Backs Bipartisan Congressional Push to Tighten China Chip Export Limits; Nvidia Loses Key Political Influence
1 week ago
According to a November 12 Bloomberg report, U.S. artificial intelligence company OpenAI is backing a new round of bipartisan legislation in Congress to tighten semiconductor export controls on China, aimed at further restricting China's access to advanced AI chips and semiconductor manufacturing equipment. OpenAI had previously taken a relatively cautious stance on related policy; its explicit support for stronger restrictions now reflects divisions within the U.S. AI industry over technology policy toward China. With Nvidia founder Jensen Huang resigning as CEO amid a series of controversies over personal relationships, export control measures that Nvidia had strongly opposed are once again gaining bipartisan momentum in Congress.
Congress is currently debating three major bills. The MATCH Act focuses on advanced semiconductor manufacturing equipment, seeking to restrict China's access to critical equipment produced by the United States and its allies and to push countries such as the Netherlands and Japan to strengthen export control cooperation. The AI Overwatch Act would require Congress to tighten oversight of the Commerce Department's approval of advanced AI chip exports, preventing the executive branch from loosening restrictions excessively. The Chip Security Act would require advanced AI chips to have location verification and tracking capabilities to reduce the possibility of transshipment to China through third countries. These bills target chip manufacturing capacity, export approval authority, and smuggling loopholes respectively, attempting to build a more comprehensive technology blockade system.
OpenAI and Nvidia have clearly divergent commercial interests on this issue. Nvidia, as the world's largest AI chip supplier, wants to continue selling products to China. Huang has warned repeatedly in the past that excessive export restrictions could push Chinese companies to accelerate development of domestic chips, ultimately weakening the global competitiveness of U.S. technology companies. OpenAI is a major user of advanced computing power, needs a continuous supply of large quantities of high-performance GPUs, and also faces competition from Chinese AI companies such as DeepSeek. Restricting Chinese companies' access to advanced chips helps maintain the computing power advantage of U.S. AI companies. Thus, although both companies are core players in the U.S. AI industry, their interests are not aligned on whether advanced chips should continue to be sold to China.
What is especially noteworthy about this policy shift is the change in Washington's political dynamics. Huang had previously become an important voice against excessively tightening chip export restrictions, relying on Nvidia's position in the global AI industry and his close contacts with the White House and members of Congress. Now, with his sudden departure, Nvidia's political influence faces realignment, while OpenAI has chosen to publicly support the bipartisan restrictive measures in Congress. However, stricter export controls may not be able to block the development of China's AI industry in the long run; they may instead further push China to expand investment in domestic GPUs, semiconductor equipment, and an independent AI ecosystem. For the United States, how to balance maintaining technological advantage, protecting the commercial interests of its own semiconductor companies, and preventing China from achieving technological self-reliance will remain the main controversy in future technology policy toward China.

From an industry-distribution perspective, the impact is not entirely uniform either. Companies focused on cloud inference and model training care more about compute costs, while makers of edge chips and devices are more worried about losing market share, so the lobbying priorities of these two types of firms may not align.