Chinese Bot Farm Used ChatGPT to Generate Anti-Data-CCenter Influence Content, OpenAI Later Identified as Inauthentic

X's Safety team uncovered 200,000 inauthentic Chinese accounts, roughly 200 of which posted AI-generated cartoons and talking points opposing U.S. data centers. Sample posts matched illustrations in OpenAI's June 2026 threat report, which identified China-based ChatGPT users producing the content — a case where a commercial AI lab's own tools generated influence material the lab later flagged as inauthentic.