OpenAI reports that accounts associated with China utilized ChatGPT to support opposition against US data centers.

      OpenAI discovered ChatGPT accounts connected to China attempting to incite local resistance against US data centres by masquerading as Americans and disseminating AI-generated material concerning increasing electricity costs. This initiative, labeled “Data Center Bandwagon” by the company, is believed to be linked to a private Chinese tech firm working for provincial government clients. However, the impact of the posts was minimal.

      These accounts utilized ChatGPT to produce English-language social media updates and cartoon graphics that portrayed executives and robots carrying bags of money while “ordinary people” shouldered the financial burden of the AI sector. Additionally, OpenAI pointed out a second initiative, “Tech and Tariffs,” that generated content critical of Trump’s tariffs and the US pursuit of global tech supremacy.

      Ben Nimmo, OpenAI's principal investigator in intelligence and investigations, clarified, “I want to be very clear: This was not an influence operation creating a debate. The debate was already there. This was a Chinese influence operation attempting to intervene in it.”

      The discourse is genuine. In 2025, local groups thwarted or postponed numerous US data centre projects that represented over $150 billion in potential investments, as reported by Data Center Watch. Senator Bernie Sanders has advocated for moratoriums on new facilities, with communities responding to overwhelmed power grids, escalating electricity costs, and pressure on water supplies.

      OpenAI noted that this campaign bears similarities to earlier operations linked to China, as identified by the Australian Strategic Policy Institute and Google’s Mandiant, which targeted firms working to decrease reliance on China in the rare earths sector. The company cautioned that such themes are “likely to remain attractive” for Chinese influence efforts, as they can be “inserted into legitimate public discussions while pushing audiences toward distrust of US institutions.”

      Several Republican lawmakers sent a letter to the Trump administration this month expressing concerns over alleged “foreign influence campaigns aimed at hindering American AI progress.” Some tech industry representatives have also promoted this narrative. However, independent researchers take a more cautious stance. Darren Linvill, co-lead of Clemson University’s Media Forensics Hub, told NPR that his team found minimal evidence of coordinated Chinese actions, stating, “We haven’t discovered much.”

      The timing aligns well for OpenAI, which is vigorously advocating for data centre development to meet the rising demand for its offerings. The company argues that AI infrastructure is critical for national competitiveness with China. Positioning domestic opposition as partly influenced from abroad supports this agenda, even if the influence operation identified by OpenAI was limited and ineffective.

      The fundamental tension remains, irrespective of who amplifies it. The energy demands of AI are escalating more swiftly than power grids can accommodate. Communities hosting data centres are incurring costs that primarily benefit tech firms located elsewhere. The fact that a few Chinese social media accounts attempted to leverage this discontent does not alter the legitimacy of the grievance itself.