In a significant development for the global semiconductor supply chain, China has agreed to address United States concerns regarding shortages of critical rare earth minerals, following the high-stakes summit between U.S. President Donald Trump and Chinese President Xi Jinping in Beijing. The agreement, detailed in a White House factsheet released on May 17, 2026, signals a potential easing of export controls that have threatened the production of advanced AI hardware and aerospace technologies.
The two-day summit, which concluded on Friday, May 16, produced a wider package of trade understandings, including Chinese commitments to purchase at least $17 billion worth of American agricultural products annually through 2028 and 200 Boeing aircraft. However, for the technology sector, the most critical outcome was the specific commitment regarding strategic minerals.
According to the White House factsheet, “China will address U.S. concerns regarding supply chain shortages related to rare earths and other critical minerals, including yttrium, scandium, neodymium, and indium.”
This announcement comes as a relief to American and allied technology manufacturers who have been grappling with the fallout from Beijing’s export controls. As reported by Reuters and Firstpost, China introduced strict export restrictions on several specialty rare earths in April 2025, in direct retaliation for President Trump’s “Liberation Day” tariffs.
The AI Semiconductor Connection
While rare earths are often discussed in the context of electric vehicles and defense systems, the specific minerals highlighted in the White House factsheet (yttrium, scandium, and indium) are deeply intertwined with advanced semiconductor manufacturing and the broader AI hardware ecosystem.
Yttrium and scandium are critical components in the production of specialized alloys and ceramics used in the manufacturing equipment that produces advanced microchips. Furthermore, indium is increasingly vital for next-generation computing architectures. For example, advanced quantum computing prototypes and highly specialized AI accelerators utilize indium for connectors between wafers and in the development of indium arsenide components. Neodymium, also mentioned in the agreement, is essential for the permanent magnets used in the cooling systems of massive AI data centers and in the motors of robotics systems.
The concentration of global supply chains in China has long alarmed Western governments. China refines over 90 percent of the world’s rare earths and has spent decades building dominance across mining, refining, and processing operations. Much of its expertise and proprietary technology remains tightly guarded and generally inaccessible to foreign firms.
The White House factsheet noted that Beijing would also address U.S. concerns over China’s export restrictions on rare earth processing equipment and technology, potentially opening the door for Western firms to improve their own domestic refining capabilities.
China’s Rare Earth Dominance as Strategic Leverage
Beijing’s control over rare earth supply chains did not happen by accident. China has spent decades systematically building its dominance across every stage of the rare earth value chain, from mining and refining to the development of proprietary processing technologies. This comprehensive control has transformed rare earths from an industrial commodity into a potent geopolitical instrument.
The April 2025 export controls were a calculated escalation, demonstrating that Beijing was willing to weaponize its resource dominance in response to American trade pressure. The controls targeted specialty rare earths that are particularly difficult to substitute or source elsewhere in the short term, maximizing the pressure on American defense contractors, semiconductor manufacturers, and clean energy companies.
For the AI industry specifically, the disruption has been felt most acutely in the supply chains for advanced semiconductor manufacturing equipment. The specialized alloys and ceramics that require yttrium and scandium are used in the components of the machines that etch the nanoscale circuits onto silicon wafers. Any disruption to this supply can create bottlenecks that ripple through the entire global chip production ecosystem, ultimately constraining the supply of the AI accelerators that are in such high demand.
A Fragile Truce in the Tech War
Despite the positive signals from Washington, the durability of this rare earth truce remains uncertain. Notably, China’s Ministry of Commerce did not mention rare earths in its official summary of the summit, released on Saturday, highlighting persistent gaps in how both governments are portraying the meeting’s outcomes.
The omission by Beijing suggests that while an understanding may have been reached, the actual implementation of relaxed export controls could be slow or conditional on broader trade concessions. This mirrors a previous agreement in October 2025, where the White House stated China had agreed to allow shipments to freely flow, yet exports of key rare earth materials remained tightly controlled in practice.
The rare earth agreement must be viewed within the broader context of the escalating geopolitical race for AI supremacy. While the U.S. has aggressively utilized export controls to restrict China’s access to advanced AI chips (like Nvidia’s H200) and the extreme ultraviolet (EUV) lithography machines needed to make them, China has leveraged its dominance over the raw materials required to manufacture those same technologies.
This dynamic has forced both nations to accelerate their push for self-sufficiency. As the U.S. attempts to rebuild its domestic semiconductor manufacturing base, China is supporting domestic memory chip champions like CXMT.
The Trump-Xi summit concluded with both leaders agreeing to meet again in the United States in September 2026. Until then, the technology industry will be closely monitoring export data to see if the promised flow of yttrium, scandium, and indium materializes, providing the critical raw materials needed to sustain the breakneck pace of global AI hardware development.