A suspected influence operation dubbed 'Data Center Bandwagon' used ChatGPT to generate anti-data center content aimed at stoking opposition to American AI infrastructure
Someone in China decided the best way to slow down American AI dominance was to make Americans hate their own data centers. It didn’t really work, but the attempt itself tells you everything about where the US-China tech rivalry is headed.
OpenAI published a threat report on June 10 detailing what it calls the “Data Center Bandwagon,” a suspected influence operation linked to the People’s Republic of China that used ChatGPT accounts to generate social media content designed to provoke opposition to US data center development. The company traced the accounts back to a private Chinese tech firm reportedly operating under contract with provincial governments.
How the operation worked
The mechanics were straightforward, if a bit ironic. Operators prompted ChatGPT in Simplified Chinese, instructing it to produce anti-data center content in both English and Chinese. They then attempted to distribute this content across platforms like X and YouTube, masquerading as American voices concerned about rising electricity costs tied to AI infrastructure.
The campaign’s core strategy was to amplify existing anxieties about energy consumption. Data centers are power-hungry beasts, and communities across the US have raised legitimate concerns about their impact on local electricity grids and costs. The operation didn’t invent these worries. It tried to pour gasoline on a fire that was already smoldering.
OpenAI noted that debates over data center operations existed well before this campaign surfaced. The influence effort was essentially attempting to exploit and stress-test narratives that were already circulating in American public discourse, probing whether foreign actors could meaningfully shift the conversation around US leadership in AI technology.
The answer, at least this time, appears to be no. The operation achieved what OpenAI described as negligible traction. No significant online amplification materialized. No policy changes resulted. The accounts have since been banned.
A pattern, not an anomaly
Here’s the thing: this isn’t the first time OpenAI has flagged China-linked operations abusing its platform. Previous threat reports from the company have identified similar influence campaigns, malware development, and even the creation of surveillance tools using its models. The Data Center Bandwagon is just the latest entry in a growing catalog of attempts.
