Mike Johnson calls opposition to AI data centers a ‘Chinese psyop,’ but polls show most Americans oppose them in their neighborhoods

House Speaker Mike Johnson on Monday described opposition to new artificial intelligence data centers as a “Chinese psyop.” The Louisiana Republican said the country should not pause the industry or pile on new rules, arguing that China could otherwise take the lead. He also said technology companies should largely police themselves.

According to Newsweek, Johnson said, “We do not need a moratorium. We do not need to jump in and hyper regulate this because we’ll lose the race to China.” He added, “The innovation is important to continue.” He later said, “That is a Chinese psyop. I really believe that. That’s not a conspiracy theory.”

Those comments arrived days after President Donald Trump hosted Chinese President Xi Jinping at the White House. The visit included a military arrival ceremony, a state dinner, a flyover, and other formal events, the report said.

Surveys show most voters reject nearby data-center projects

Recent polling has shown large shares of Americans uneasy about new facilities in their own communities, across party lines, and in urban, suburban, and rural areas. A September Pew Research Center survey found that 60 percent of Americans said they would be “not too” or “not at all” comfortable with a new data center in their community. Twenty-six percent said they would be extremely, very, or somewhat comfortable.

In that Pew survey, 54 percent of Republicans and Republican-leaning independents said they would be uncomfortable, compared with 68 percent of Democrats and Democratic-leaning independents. Fifty-four percent of conservatives expressed discomfort, while 81 percent of liberals did.

A University of Massachusetts Amherst national poll released this month found that 65 percent of Americans opposed building an AI data center in their community, versus 11 percent who supported it. Among Republicans, 52 percent opposed construction and 18 percent supported it, the poll showed.

A Reuters/Ipsos survey this month found that 73 percent of Americans were concerned AI companies were not doing enough to prevent potentially catastrophic consequences. A majority favored federal oversight of AI safety standards, and 55 percent supported slowing AI development. Respondents prioritized safe and responsible development over global competitiveness by a substantial margin, the survey found.

Johnson’s stance has clashed with other Republicans who have questioned electricity prices, grid strain, and local impacts. At a Washington event earlier this month, former Trump adviser Steve Bannon and independent Sen. Bernie Sanders were among those calling for stronger AI guardrails. Rep. Chip Roy also raised concerns about the effects on communities. 

In Michigan, Rep. Tom Barrett has opposed taxpayer subsidies for data centers while campaigning with Johnson. House Republicans have been preparing the Ratepayer Protection Act, which would have state regulators consider requiring data centers to cover some costs of new power generation.

X said in August that its safety team identified a suspected Chinese bot farm of about 200,000 accounts. The debate over data centers has become a problem for both parties, as MAGA and progressives are now on the same side.

Roughly 200 of those accounts posted material that could shape the U.S. debate over AI and energy, including claims that data centers were raising electricity prices and straining the grid, X stated. The report said that findings offer some support for Johnson’s foreign-influence concern but do not show that people who oppose data centers are being manipulated by China.

Nvidia has announced an Open Agent Safety Platform meant to tighten controls on autonomous AI agents. The company said the system can quarantine an agent that tries to leave its limits within milliseconds. OpenShell sets what files, networks, and processes an agent may use, while Sentry runs separately on Nvidia’s BlueField-4 data processing units.

CEO Jensen Huang has said AI safety needs “full-stack engineering” from the model through software and hardware. Nvidia said more than 100 organizations are taking part in or backing the effort.