OpenAI says fake China accounts spread AI‑generated posts about data center risks

OpenAI has reported that groups likely operating from China used ChatGPT to plan covert influence campaigns targeting Americans’ opinions about AI data centers. One cluster, dubbed “Data Center Bandwagon,” generated English‑language posts and images portraying data centers as drivers of rising electricity costs, while posing as ordinary Americans online. Another cluster produced multilingual content criticizing US tariffs and tech policy, aiming at audiences including Taiwanese users. Although the campaigns failed to gain meaningful engagement, OpenAI says they are significant because operators attempted to secretly insert themselves into an ongoing US debate while hiding their identity and motives. The campaigns also targeted Chinese dissidents abroad with harassment content generated by ChatGPT. OpenAI notes it cannot determine why the groups chose an American chatbot instead of Chinese alternatives.

Read the full story on Engadget →