X Finds 200,000-Account Bot Farm in Suspected Chinese AI Policy Influence Push
The platform removed accounts that broke its rules, but the content it identified landed in a U.S. fight over power costs, grid capacity and local development that is already politically charged.
Listen to this story
The audio brief
Story brief
3 key pointsX’s investigation adds an influence-operation dimension to the U.S. fight over AI infrastructure: roughly 200,000 suspected Chinese inauthentic accounts included about 200 that promoted anti-data-center narratives. X suspended accounts violating its Authenticity policy but did not identify a specific Chinese actor. The activity landed amid genuine opposition to electricity demand, local construction, and rising...
- 01
Only about 200 accounts in the roughly 200,000-account network were identified as pushing AI and energy-policy narratives.
- 02
Posts alleged data centers raise household electricity prices and strain grids, sometimes using AI-generated cartoons.
- 03
An Economist/YouGov poll found 47% viewed new data-center construction negatively, versus 24% positively.
X says it found an approximately 200,000-account bot farm tied to suspected Chinese inauthentic accounts. Within it, 200 accounts posted material that could manipulate debate over American AI and energy policy, putting foreign influence into a consequential infrastructure fight.
The content focused on power and cost
X’s Safety team investigated suspected Chinese inauthentic accounts involved in influence operations and identified the bot farm. It said a small subset posted in a way that could manipulate a legitimate debate about U.S. AI and energy policy.
The posts claimed that AI data centers were raising household electricity prices and straining the power grid. X said some used AI-generated cartoons portraying data-center operators as profiting at the public’s expense; others warned of higher energy rates and showed corporations destroying communities.
X said it suspended accounts that violated its Authenticity policy. Its finding identifies suspected Chinese inauthentic accounts, rather than assigning the operation to a named actor.
An Economist/YouGov poll found that 24% of Americans held this view.
The same poll found a 23-point gap between negative and positive views.
The campaign met an existing political conflict
Public resistance is visible at the project level, too. A Deseret News/Hinckley Institute poll found 60% of Utah voters opposed the proposed Stratos Project data center, while 26% supported it.
The scale of the power question is substantial: the largest proposed facilities for training AI models could require up to 10 gigawatts of electricity, according to Deseret News. New York and Pennsylvania have restricted data-center construction.
A second China-linked allegation
OpenAI said in June that it banned a group of accounts likely originating in China that targeted U.S. audiences in a covert influence operation. That group claimed data centers make life more expensive.
X’s removals address rule-breaking accounts, not the policy choices facing communities. The next test is whether debate over new facilities can distinguish coordinated manipulation from the real questions raised about electricity demand and construction.
Sources
- foxbusiness.comX reveals foreign adversary behind fake accounts disseminating anti-data center messaging
- deseret.comThe campaign to make Americans hate data centers
