PRC-linked influence operations are targeting AI debates in the US

Our mission at OpenAI is to ensure artificial general intelligence benefits all humanity. We achieve this by deploying innovations to build democratic AI, governed by common-sense rules, helping solve complex problems while protecting against harm. This also involves identifying and disrupting attempts by authoritarian regimes to use AI systems for coercion, surveillance, or covert interference in democratic societies.

We banned two clusters of ChatGPT accounts, likely originating from China, for using our models in covert influence operations. These operations promoted narratives to manipulate legitimate debates about American AI and broader tech policies. This interference underscores the evolving landscape of digital threats.

The first cluster, termed the "Data Center Bandwagon" campaign, generated social media comments and images. These falsely claimed that AI data center expansions were raising electricity prices for families. The second cluster, the "Tech and Tariffs" campaign, criticized US tariffs as attempts to dominate technological competition and specifically avoided mentioning China's leader, Xi Jinping, in their output.

This targeting of OpenAI and US data center infrastructure is significant. While these operations didn't appear to shift public opinion, they show PRC-origin influence operators testing narratives against critical AI infrastructure. This infrastructure is a foundation of US technological leadership, economic growth, and the broader democratic AI ecosystem.

The operators tried to exploit existing public concerns about energy prices and the local impacts of data center development. However, we found no evidence of meaningful breakout beyond their own activities. Foreign influence operations frequently latch onto local issues and deeply held beliefs to build credibility, amplify divisions, or exacerbate public distrust.

By publishing these findings, we aim to help partners—industry, governments, and civil society—better identify and disrupt foreign threat actors. This proactive approach is crucial to counter attempts to manipulate public debates, weaken democratic institutions, and advance totalitarianism using AI for surveillance, censorship, and control.