AI company OpenAI has uncovered two clusters of ChatGPT accounts, believed to have originated from China, that were involved in covert influence operations. These clusters aimed to manipulate discussions on American AI and broader tech policies. The first cluster, named the “Data Center Bandwagon” campaign by OpenAI, spread misinformation about data center buildouts for AI leading to increased electricity prices for families.
The second cluster focused on criticizing US tariffs as a means to dominate technological competition. Notably, this cluster avoided mentioning China’s leader Xi Jinping and instead emphasized President Donald Trump. OpenAI highlighted that this cluster was linked to a network of likely fake social media accounts that falsely claimed ChatGPT user data had been compromised, termed as the ‘Tech and Tariffs’ campaign.
OpenAI emphasized the significance of these operations in testing narratives against AI infrastructure, a crucial aspect of US technological leadership. While these operations attempted to exploit public concerns about energy prices and data center development impacts, they did not achieve significant traction beyond their own activities. OpenAI stressed that these operations aimed to interfere in ongoing American debates on AI capabilities while concealing their identities and motives.
The company’s objective is to assist various stakeholders in identifying and thwarting foreign attempts to manipulate public debates, undermine democratic institutions, and promote totalitarianism with AI characteristics, such as using AI for surveillance and control over social, political, and private spheres.
