OpenAI revealed a troubling influence operation where China-linked actors weaponized ChatGPT to shape American opinions about AI infrastructure. The company identified two clusters of fake accounts that generated anti-data center messaging, including talking points about rising electricity costs and comic strips highlighting consumer bill increases.

The first group, labeled the "Data Center Bandwagon," posed as Americans across social media platforms while sharing AI-generated content about power grid strain. These accounts specifically targeted communities near data centers, using real news links about electricity auctions alongside fabricated narratives. A Bloomberg report noted power costs have surged up to 267 percent in areas surrounding these facilities, making the messaging resonate with genuine concerns.

A second cluster focused on criticizing US tariffs and tech policies, deliberately avoiding references to Chinese leadership while crafting content in multiple languages for Taiwanese audiences. Both groups uploaded strategic documents to ChatGPT outlining their objectives and methods for creating undetectable fake accounts.

Despite the sophisticated approach, OpenAI noted these campaigns achieved minimal authentic engagement and failed to significantly shift public opinion. However, the operations represent a concerning attempt to covertly insert foreign influence into domestic debates about America's AI future, exploiting legitimate issues for strategic advantage.