OpenAI reports it banned accounts linked to China for allegedly using ChatGPT to influence US public opinion on AI data centre expansion and energy costs.
OpenAI has said it identified and banned a group of accounts it believes were linked to China for attempting to run covert influence operations using ChatGPT, targeting public debate around artificial intelligence infrastructure in the United States.
In a report released this week, the company said the accounts were used to generate social media content aimed at shaping public opinion on AI data centres, particularly focusing on concerns over electricity prices and energy consumption in local communities.
The San Francisco-based company said the activity appeared designed to “manipulate a legitimate debate about American AI” by amplifying existing concerns rather than introducing entirely new narratives.
According to OpenAI, some of the content created by the accounts included images and posts suggesting that AI data centres were driving up household electricity bills. One example described by the company featured a cartoon of a businessman holding money bags while a family reacted to a rising energy bill.
A separate cluster of accounts was also found to be producing material related to US trade policy and technological competition with China. OpenAI said these posts framed tariffs as part of a broader effort to dominate global tech development and instructed that references to Chinese leadership should be avoided.
The company stated that while the campaign attempted to insert itself into ongoing political and economic discussions, there was no evidence it had a significant real-world impact on public opinion.
OpenAI noted that foreign influence operations often exploit existing social concerns to increase credibility and deepen divisions within public debate.
In response, China’s embassy in Washington said it was not familiar with the report but rejected what it called unfounded accusations against the country, reiterating opposition to any claims it described as politically motivated.
The findings come amid growing global scrutiny of how artificial intelligence tools may be used not only for productivity and communication but also for information manipulation and digital influence campaigns.
Experts say the case highlights how emerging AI systems are increasingly becoming part of broader geopolitical tensions, especially as countries compete for leadership in advanced technologies.
