NextFin News - On December 24, 2025, U.S. President Donald Trump officially authorized American semiconductor firms Nvidia and Advanced Micro Devices (AMD) to resume the sale of advanced AI chips to China, marking a significant policy shift. This decision reverses the export ban imposed during the prior administration, which restricted shipments of leading-edge AI computing chips over national security concerns. Under the new framework, Nvidia will pay a 25% fee on its H200 AI chip sales to Chinese customers, while AMD secured a 15% fee arrangement for their China-specific MI308 AI accelerators. This resumption follows President Trump's December 8, 2025 announcement and extends potential export openings to other manufacturers such as Intel.
The U.S. President disclosed on social media platform Truth Social that he informed Chinese President Xi Jinping of the export resumption, receiving a positive response. Initial shipments by Nvidia are scheduled before the Lunar New Year in mid-February 2026, utilizing existing stock estimated at 5,000 to 10,000 H200 chip modules—equivalent to roughly 40,000 to 80,000 individual chips. Concurrently, AMD’s China-focused MI308 chips, engineered to comply with U.S. export laws, are under consideration for purchase by major Chinese tech firms like Alibaba, targeting volumes between 40,000 and 50,000 units.
Chinese regulators have engaged with domestic technology giants such as Alibaba, ByteDance, and Tencent to assess demand, though Beijing is reportedly evaluating measures to mandate a mix of imported and domestically produced chips per purchase. Approval of final transactions hinges on government authorization, highlighting cautious yet pragmatic coordination between the two nations.
This pivot reflects a strategic recalibration by the Trump administration, balancing economic and national security priorities amid intensifying U.S.-China competition in artificial intelligence and semiconductor technology. The policy aims to maintain U.S. technological dominance by controlling the export process with embedded fees, thereby creating a revenue stream while regulating China’s AI hardware access.
Analyzing this development reveals multiple dimensions. First, the authorization suggests a recognition of the intertwined nature of global semiconductor supply chains, where outright bans risk isolating U.S. companies from a major market and ceding ground to non-U.S. global competitors. The imposed fees represent an innovative economic lever to offset security risks and incentivize compliance.
Empirically, the H200 chip outperforms the previously export-approved H20 chips by a factor of nearly six in computational power, equipping Chinese AI developers with unprecedented hardware capacity. Such capability could accelerate China’s AI development curve but simultaneously imposes dependencies on U.S. high-tech imports under strict regulatory scrutiny.
The strategic choice also signals a nuanced approach to technology export control: rather than blanket prohibitions, conditional access with financial and regulatory guardrails may better serve long-term interests. This model may establish a precedent for handling sensitive technology transfers amid geopolitical rivalries while sustaining market participation for American firms.
Looking forward, this policy shift is likely to influence several trends. The resumption could prompt Chinese domestic chipmakers to accelerate innovation to compete with imported Nvidia and AMD solutions, fostering a dynamic competitive environment. Conversely, it may temper China’s urgency to develop fully indigenous AI chips, given improved access to world-class hardware.
U.S. firms stand to benefit from renewed revenue streams and enhanced market presence in China, which remains one of the largest and fastest-growing AI markets globally. However, the administration must carefully monitor technology diffusion risks that could empower Chinese military or surveillance capabilities, a paramount national security concern.
On the international stage, this recalibrated export policy might encourage other allied nations to reconsider similar restrictions, influencing the global semiconductor ecosystem governance. It also underlines the increasing complexity of technology diplomacy embedded in great power competition, where economic, security, and technological factors intertwine.
In conclusion, U.S. President Trump's authorization of advanced AI chip exports to China signals a sophisticated balancing act between maintaining American technological leadership and managing geopolitical risks. The policy’s success will depend on rigorous enforcement of export controls, ongoing bilateral communication, and adaptive strategies in semiconductor innovation and supply chain resilience.
According to Dataconomy, this landmark move has stirred debate across policy, industry, and security spheres, marking a pivotal point in U.S.-China technology relations that will reverberate throughout 2026 and beyond.
Explore more exclusive insights at nextfin.ai.
