Advantage China?: Data center opposition, Jensen Huang’s resignation changing America’s AI rollout

apg-wi.com

Source published: Site published:

Original article · فارسی

Germany Nvidia

Nvidia CEO Jensen Huang gives his keynote speech at the company's GTC event in Berlin, Germany, on Wednesday, Oct. 21, 2026. (AP Photo/Markus Schreiber)

Markus Schreiber

Artificial intelligence data centers are likely to face more protests, temporary bans and rules after Democrats won big in the midterm elections.

Many Democrats campaigned on slowing the growth of AI. A survey found that 76% of all voters oppose having data centers in their local communities.

Voters across the United States approved 40 local bans and other limits on data centers.

More restrictions are expected when Democrats take control in Congress and in state governments across the country in January.

“Local opposition has become another major limit on AI expansion,” said Jock Putney, a technology executive based in San Francisco and CEO of three companies: Nuvolum, Quantum PRM and Stemodontics.

Putney said opposition to data centers can delay projects and limit the amount of computing power available for AI. This could raise prices for customers and make it harder for smaller companies in the United States to compete.

“Amazon, Microsoft, Google and SpaceX can afford a long local fight and can bring their own power,” Putney said.

A slower rollout of AI in the United States could also give China an advantage as it tries to challenge America's leading position in technology.

“For China, prolonged U.S. delays could create a competitive opening,” said Ian Arguno, head of analytics and artificial intelligence and machine learning at Realmo, a commercial real estate information company.

But Arguno said the situation has many different parts.

“Calling the election itself a Chinese victory is premature. China’s progress still depends on access to computing hardware, reliable electricity, capital and customers,” Arguno said.

‘RELATIVE GAIN’

Arnold Platz, CEO and founder of General Future Holding, a company that recycles AI technology, said China has some advantages in the global race. It is willing to build data centers without long permit delays or political opposition, and it has important resources needed to power AI.

“Every month of delay in the U.S. build-out is a relative gain for China, but China’s real advantage has nothing to do with elections; it is speed,” Platz said. “I see how fast they build power plants, factories and grid (infrastructure), and it is a different tempo from anything in the West. China also refines most of the critical minerals that end up inside every server.”

China is building large new AI data centers in Inner Mongolia and other inland areas. Earlier this year, it opened an underwater data center powered by wind near Shanghai. Other underwater data centers have been built off Hainan Island.

“China is now contemplating the idea of directing roughly $295 billion into AI data centers for five years, and 80% or more of all hardware and software will be made in China,” said Stanislav Kazanov, head of cybersecurity and data at Innowise, an international software company.

The United States has the money, most of the world's existing data centers and important technology advantages. But China has an advantage in the minerals and rare earth elements needed to build AI infrastructure. China produces or controls most of the world's copper, cobalt, lithium, gallium and nickel. All these materials are needed for data centers and technology equipment.

Natan Rosengaus, CEO of Bleeding Edge Technologies, a company based in Mexico City that designs and builds modular “AI factories” containing data centers and other technology equipment, expects China to be “very aggressive” in AI.

Rosengaus said this could push the United States to seek new AI partners if data center projects are delayed or stopped in the country.

“In that sense, we see that Mexico could become the natural market, because of location, innovative projects that can offer infrastructure engineered for hyperscale workloads, and the history of both countries [having] jointly built and operated complex infrastructure and supply chains,” Rosengaus said.

President Donald Trump has had difficult relationships with Canada and several European countries. However, his relationship with Mexican President Claudia Sheinbaum has been warmer.

“The new conditions favor the opportunity of creating a more integrated North American AI infrastructure ecosystem, where capacity can be deployed wherever energy, speed, and economics make the most sense,” Rosengaus said.

But Mexico and other Latin American markets have problems with their water and electricity systems.

Others see opportunities for data centers in Asian countries such as Malaysia, Thailand, Indonesia and Vietnam.

Kristopher Clark, a technology and infrastructure architect based in Kansas City, said he is seeing major AI construction projects in Vietnam. He is also seeing data centers being built to serve local markets in Japan and India.

“A lot of countries also want AI infrastructure inside their own borders now, for legal and political reasons. That’s new capacity being added. It isn’t capacity leaving America,” Clark said. “I’m actually in Vietnam at the moment, and the amount of data center construction going on in this part of the world is hard to miss.”

‘A DIFFERENT STORY’

Clark said “China is a different story.”

“Export controls have split it off into its own ecosystem,” Clark said, referring to U.S. rules that limit technology sales to China.

Democrats taking control of Congress are joining conservative lawmakers who take a tough position on China. Together, they are pushing for stricter U.S. limits on exports of AI chips and technology. This comes amid investigations into whether an alleged affair involving former NVIDIA CEO Jensen Huang and Singapore technology executive Alice Huang, a Chinese citizen who is not related to him, was a Chinese intelligence operation known as a “honeypot.”

The investigations are looking at whether NVIDIA chips reached Chinese buyers and whether Jensen Huang tried to persuade the Trump administration to make technology export rules involving China less strict.

Jensen Huang resigned on Thursday, Nov. 12, after reports of a second affair with an NVIDIA executive. Trump said that same day that he wants former Apple CEO Tim Cook to take over as head of NVIDIA.

Huang had also supported Trump on AI policy.

“We are giving up the lead in AI, in my view, to China. Jensen was the most optimistic supporter of AI,” said CJ Muse, senior managing director and technology industry analyst at investment firm Cantor Fitzgerald, in an interview with CNBC on Thursday.

Muse said Huang and NVIDIA created the technology used to power the most advanced AI models. Huang supported more technology trade with China. He argued that the global growth of AI should be powered by American technology and that trade restrictions reduce access to markets.

U.S. Sen. Chris Coons, a Democrat from Delaware, is concerned about Huang's connections to Alice Huang and his support for more open technology trade with China.

“NVIDIA CEO Jensen Huang has spent years lobbying Trump to allow advanced AI chip sales to China. Now, serious questions are emerging about his ties to an alleged Chinese spy and what Trump knew,” Coons said in a statement. “Congress needs answers, and Americans deserve better national security protections.”

Coons, U.S. Sen. Elizabeth Warren, a Democrat from Massachusetts, and U.S. Sen. Pete Ricketts, a Republican from Nebraska, support tighter limits on U.S. chip and technology exports to China because of national security concerns.

‘NARROWS THE GAP’

Trump compares AI to the Industrial Revolution. He says it is important for the United States to “win” the new global race in technology.

Does that mean avoiding data center construction in the United States is like refusing to build factories at the start of the Industrial Revolution?

“A U.S. slowdown narrows the gap where China is strongest,” said Sean Caricola, a partner at APAC Advantage, a Vietnam-based company that helps businesses set up operations in Asia. “But China has a real constraint when it comes to accessing top chips. Nothing in this midterm makes that any easier for them.”

Caricola believes companies in the AI industry will make more backup plans because the election results have strengthened opposition to data centers in the United States.

“To be clear, nobody is leaving America when the chips, the capital and the customers are still there. What changes is the cost of time,” he said. “When a ballot measure can kill a project, every developer will need to have a Plan B, and the Plan B is Asia. The midterms possibly just shortened that debate.”

As it does in other industries, the Chinese government supports its AI industry with money and other help. This includes machine learning and data centers. Unlike in the United States, data centers in China do not have to go through the same permit and approval process.

“I see it as an opening for China’s own AI industry. China can build power and computing infrastructure at scale, and renewable-rich western provinces give developers more options for siting new capacity,” said Aaron Henry, founder and managing director of Foundeast Asia Co. Ltd., a communications company based in Thailand.

“China’s other advantage is distribution. Open-weight models from developers such as DeepSeek and Alibaba’s Qwen team let companies download and adapt models and run them on infrastructure they choose,” Henry said.

“That gives Chinese models a route to wider adoption even when the computing happens outside China.”

Mike Sunnucks is a national reporter for Adams MultiMedia. He writes about business, politics and technology.