Microsoft Urges Trump Administration to Abandon Planned Ban on Chinese Data Centre Components

Microsoft is urging the Trump administration to abandon plans to restrict imports of Chinese-made components used in artificial intelligence (AI) data centres, warning that the proposed restrictions could disrupt construction projects and undermine the United States’ position in the global AI race.

According to a report by Reuters, Microsoft is among several major technology companies and industry organisations lobbying US officials against the proposed restrictions, which would target Chinese-made optical transceivers.

The components play an important role in modern data centres, allowing servers, switches and other networking equipment to communicate through high-speed fibre-optic connections. They are particularly important in large-scale AI facilities, where enormous volumes of data must move between thousands of processors and networking devices.

The US government has been considering restrictions on new models of Chinese optical transceivers amid concerns that equipment manufactured by Chinese companies could potentially be exploited to steal information, introduce malware or disrupt services operating within American data centres.

The Federal Communications Commission (FCC) is reportedly working on the proposed restrictions as part of broader efforts to reduce reliance on Chinese technology in critical infrastructure.

However, Microsoft and other technology companies believe that introducing a ban could create significant supply chain problems at a time when the United States is investing heavily in expanding its AI infrastructure.

Microsoft warns of delays to AI infrastructure

Microsoft’s concerns are partly driven by the scale of its own data centre expansion programme. According to market intelligence company Cleanview, the company has 35 data centres under construction across the United States, with another 134 planned.

The company reportedly met with officials from the White House Office of Science and Technology Policy and the National Security Council in September to explain the potential supply chain difficulties associated with restricting Chinese optical transceivers.

Industry representatives argue that data centre operators have already invested heavily in AI processors and other infrastructure. Without sufficient optical networking equipment, some of that expensive hardware could remain unused while companies attempt to secure alternative components.

John Miller, general counsel at the Information Technology Industry Council (ITI), warned Reuters that the proposed restrictions could seriously disrupt data centre construction.

Some large data centres require millions of optical transceivers to connect servers, switches and routers. Finding alternative suppliers capable of delivering comparable quantities could prove difficult.

China dominates the optical transceiver market

One of the biggest challenges facing the US technology industry is its dependence on Chinese manufacturers.

According to Counterpoint Research, China accounts for approximately two-thirds of global optical transceiver production. Chinese manufacturers, including Zhongji Innolight and Eoptolink, are major suppliers to the global data centre industry.

Zhongji Innolight alone accounts for approximately 27 per cent of the global data centre transceiver market.

Although American companies including Coherent and Lumentum manufacture competing optical networking products, their current production capacity is not considered sufficient to replace Chinese suppliers quickly.

A sudden restriction could therefore increase costs, create shortages and delay the deployment of new AI infrastructure.

The ITI has previously written to the FCC urging the regulator not to add optical transceivers to its trade blacklist, warning that supply disruptions could damage American AI development and competitiveness.

National security concerns remain

The proposed restrictions form part of a wider effort by Washington to limit the use of Chinese technology in critical American infrastructure.

US officials have expressed concerns that Chinese-made networking equipment could provide opportunities for espionage, data theft or disruption of essential services.

These concerns have become increasingly important as data centres supporting AI services, cloud computing and other critical applications become central to the US economy.

The debate has also attracted attention from US lawmakers. In September, a bipartisan group of senators introduced legislation seeking to prevent Chinese-made optical transceivers from being used in sensitive federal government systems.

The proposed legislation specifically targets equipment from manufacturers including Eoptolink and Zhongji Innolight.

However, the approach being considered by the FCC could be broader, potentially restricting imports of new Chinese optical transceiver models.

There have also been discussions about potentially targeting particularly advanced equipment capable of handling 3.2 terabits of data per second. Such technology is not expected to reach large-scale production until 2028, potentially giving American manufacturers additional time to expand their manufacturing capabilities.

A difficult balancing act for Washington

The situation highlights the growing tension between national security concerns and the commercial realities of building large-scale AI infrastructure.

The Trump administration wants to reduce American dependence on Chinese technology, particularly in areas considered strategically important. However, restricting access to components that currently dominate global supply could make it more difficult and expensive for American technology companies to expand their infrastructure.

Microsoft and other industry representatives believe that disrupting the supply of optical transceivers could ultimately weaken the United States’ competitive position against China in artificial intelligence.

For Washington, the challenge will be finding a way to address potential security risks without creating additional obstacles for an industry already facing enormous demand for computing capacity.

The FCC has not publicly confirmed its final position on the proposed restrictions, and Microsoft has declined to comment on its discussions with US officials.

The outcome could have significant implications for the global data centre supply chain, particularly as technology companies continue investing billions of dollars in the infrastructure needed to support the next generation of AI services.