OpenAI has identified and banned China-linked accounts that used its ChatGPT tool to orchestrate social media campaigns aimed at influencing U.S. debates over tariffs and AI data centers. The campaigns, though ineffective, highlight ongoing efforts by pro-China actors to exploit American political divisions.
Two Targeted Operations
OpenAI uncovered two distinct operations that utilized ChatGPT to create content, including posts, comments, and political cartoons. One campaign, named 'Data Center Bandwagon,' focused on claims that AI data centers were increasing electricity costs for American households. The second, dubbed 'Tech and Tariffs,' criticized former President Donald Trump's tariff policies and the U.S. push for global tech dominance.
'This was not a case of an influence operation creating a debate. The debate existed already. This was an influence operation from China trying to interfere in it,' said Ben Nimmo, principal investigator on OpenAI's intelligence and investigations team.
Campaign Tactics
The 'Data Center Bandwagon' campaign involved ChatGPT-generated comic strips about power grid capacity and electricity prices, posted on social media alongside legitimate news articles. Another operation produced political cartoons depicting Trump advocating for 'America First' policies while undermining global cooperation in technology.
Limited Impact
Despite their efforts, the campaigns failed to gain significant traction online. OpenAI noted that this marks the first instance of China-linked actors using its AI models to meddle in U.S. tech policy debates. The company emphasized that such operations are likely to evolve as foreign actors increasingly leverage AI tools to scale their influence efforts.
This revelation underscores the growing threat of AI-driven foreign interference in U.S. political and economic debates, raising concerns about the need for robust safeguards against such exploitation.