X Uncovers China-Linked Bot Network Posting Anti-AI Data Center Content

X says it has uncovered a large network of suspected inauthentic accounts, including hundreds of accounts that promoted messages criticizing the growing number of AI data centers in the United States.

According to X’s Safety team, the investigation identified a network involving approximately 200,000 accounts suspected of being connected to China. The company said around 200 of those accounts were specifically involved in posting content about AI data centers, electricity prices, and pressure on local power grids.

X said the activity appeared designed to influence legitimate discussions surrounding AI, energy consumption, and US technology policy.

AI Data Centers Become the Focus of the Campaign

The accounts reportedly shared comic strips, images, and written posts criticizing the impact of large AI data centers on communities.

Some of the posts claimed that data centers could contribute to higher household electricity costs and place additional pressure on local power grids.

The campaign also included AI-generated illustrations portraying data center operators as businesses benefiting financially while ordinary households were left to deal with rising energy costs.

X did not publicly identify the accounts involved in the operation. However, the company shared examples of some of the content as part of its announcement.

The platform said it remains focused on maintaining an authentic environment for discussions about important public issues and will suspend accounts that violate its authenticity policies.

ChatGPT Was Reportedly Used to Create the Content

One of the more notable aspects of the investigation is the reported use of AI to produce material for the campaign.

X said examples of the content appeared to match material previously identified by OpenAI in a report published in June.

According to OpenAI, users believed to be operating from China had used ChatGPT to generate illustrations and talking points related to opposition to AI data centers.

The content was then reportedly distributed through social media accounts.

This highlights an increasingly important issue in the AI era: the same tools that can help people create legitimate content can also be used to produce large volumes of material for coordinated influence campaigns.

Real Concerns Can Still Be Part of an Inauthentic Campaign

The situation is particularly complicated because the underlying topic is not necessarily fictional.

AI data centers consume significant amounts of electricity, and their rapid expansion has raised legitimate questions about energy infrastructure and costs.

Communities located near large computing facilities have increasingly debated how these projects could affect electricity demand, local infrastructure, and utility costs.

According to figures cited in the original reporting, Goldman Sachs estimated that electricity prices had increased 6.9% year over year as of February 2026 and expects demand from AI facilities to continue putting pressure on electricity markets.

That means criticism of AI data centers should not automatically be considered disinformation.

The concern raised by X is instead about the possibility that coordinated, inauthentic accounts were being used to manipulate an otherwise legitimate public debate.

OpenAI Previously Reported Similar Activity

X’s findings appear to overlap with activity previously documented by OpenAI.

In its June report, OpenAI said accounts believed to be based in China had used ChatGPT to create content related to AI data centers and then distribute it through social media.

OpenAI also reported that some of the same users used ChatGPT for other political content, including insults targeting Chinese dissidents and political commentators, as well as criticism of US tariffs and technology policies.

The activity demonstrates how generative AI can potentially be incorporated into broader influence operations.

Instead of manually creating every image or piece of text, operators can use AI tools to produce content more quickly and at a larger scale.

A Bigger Challenge for Social Media Platforms

The incident also highlights the difficulty platforms face when trying to distinguish genuine political or economic discussions from coordinated manipulation.

A post criticizing AI data centers may look completely legitimate on its own.

The problem becomes more apparent when large numbers of seemingly independent accounts begin sharing similar messages, images, or arguments.

AI can make this problem more difficult because content can be produced quickly and adapted for different audiences.

For social media companies, detecting coordinated behavior may therefore require looking beyond individual posts and examining account activity, relationships between accounts, posting patterns, and other signals.

AI Is Becoming Part of the Influence-Operation Toolkit

The reported campaign is another example of how generative AI is becoming part of the wider information ecosystem.

AI tools can be used for legitimate journalism, research, education, marketing, and creative work. At the same time, they can also make it easier to generate misleading or coordinated content.

This does not mean that AI-generated content is automatically malicious. The important issue is how the technology is being used and whether people are being misled about the origin or authenticity of the information they encounter.

For platforms such as X, identifying coordinated inauthentic behavior could become increasingly important as generative AI makes content production faster and cheaper.

The growing use of AI in online campaigns also highlights the broader security challenges created by increasingly capable artificial intelligence. Businesses are already looking at new ways to protect AI systems and data from emerging threats. For more information, see our article How AI Is Changing Cybersecurity.

What This Means for AI and Online Information

The case demonstrates a broader challenge that is likely to become more important as AI adoption grows.

Public debates about energy, technology, and government policy can involve genuine disagreements. But when automated systems and coordinated networks are used to artificially amplify particular viewpoints, it becomes harder for users to determine how representative those viewpoints actually are.

For users, this makes it increasingly important to evaluate online information carefully, particularly when large numbers of accounts appear to be pushing nearly identical narratives.

For technology companies, meanwhile, the challenge is finding ways to protect platforms from manipulation without suppressing legitimate debate.

Our Thoughts

X’s discovery of the suspected China-linked network shows how AI and social media manipulation are becoming increasingly connected.

According to X, hundreds of accounts within a much larger network were involved in promoting anti-AI data center content, while OpenAI had previously reported similar activity involving ChatGPT.

The controversy also illustrates an important distinction: concerns about AI data centers, electricity demand, and energy prices can be legitimate, but coordinated activity designed to artificially influence public opinion presents a separate problem.

As generative AI becomes more capable, platforms and users will need to become increasingly careful about understanding not only what information is being shared, but also who is behind it and how it was created.

Source: Engadget