OpenAI has disrupted two covert influence operations originating from China that utilized its flagship ChatGPT models to generate content aimed at manipulating American public discourse. The operations, detailed in a threat intelligence report released by OpenAI on June 10, 2026, specifically targeted sensitive domestic debates surrounding the construction of artificial intelligence data centers and the imposition of US tariffs.
The disclosure highlights the evolving nature of state-linked disinformation campaigns, demonstrating how foreign actors are increasingly leveraging advanced generative AI to produce localized, context-aware propaganda at scale. While OpenAI noted that the campaigns failed to achieve significant traction, the targeting of critical AI infrastructure marks a notable escalation in the US-China technological rivalry.
The “Data Center Bandwagon” Campaign
The first cluster of banned accounts, which OpenAI dubbed the “Data Center Bandwagon” campaign, focused on exploiting local anxieties regarding the environmental and economic impact of AI infrastructure. The operators used ChatGPT to generate social media comments and images claiming that the massive buildout of data centers required to train and run AI models was directly responsible for increasing electricity prices for average American families.
This narrative is particularly potent given the genuine, ongoing debate within the United States regarding the immense energy consumption of AI facilities. By latching onto a legitimate domestic concern, the operators attempted to build credibility and amplify existing divisions.
OpenAI’s report emphasized the strategic significance of this targeting: “The targeting of OpenAI and US data center buildouts is significant not because the operation appears to have shifted public opinion, but because it shows PRC-origin influence operators testing narratives against AI infrastructure – a foundation of US technological leadership, economic growth and the broader democratic AI ecosystem.”
The attempt to stoke opposition to data center construction aligns with broader geopolitical objectives. As the US and China vie for dominance in artificial intelligence, the physical infrastructure that powers these models has become a critical strategic asset. Disrupting or delaying the expansion of US computing capacity through covert influence operations represents a novel vector in the ongoing tech war.
The “Tech and Tariffs” Campaign
The second disrupted cluster, named the “Tech and Tariffs” campaign, focused on international trade policy. These operators used ChatGPT to generate content criticizing US tariffs, framing them as aggressive attempts to dominate technological competition rather than legitimate economic measures.
Interestingly, OpenAI’s analysis of the prompts used by these operators revealed specific instructions regarding the content’s framing. The operators explicitly directed the AI models not to include mentions of Chinese leader Xi Jinping in the generated output, instructing the system to focus solely on criticizing US President Donald Trump and his administration’s trade policies.
This campaign was also connected to a network of likely inauthentic social media accounts that sought to undermine OpenAI’s credibility directly. These accounts propagated false claims that ChatGPT user data had been compromised, an allegation that OpenAI stated was entirely fabricated. The dual-pronged approach, attacking US policy while simultaneously attempting to discredit the leading American AI developer, demonstrates a coordinated effort to weaken the US position in the global tech landscape.
The Efficacy and Future of AI Disinformation
Despite the sophisticated use of generative AI to craft these narratives, OpenAI concluded that the operations were largely ineffective. The company stated that it “found no evidence of meaningful breakout beyond its own activity,” indicating that the generated content failed to go viral or significantly influence genuine public debate.
However, the discovery of these campaigns serves as a stark warning about the future trajectory of information warfare. Foreign influence operations have historically relied on human operators to manually craft and disseminate propaganda, a process that is often slow and easily identifiable due to linguistic or cultural inaccuracies. Generative AI tools like ChatGPT drastically lower the barrier to entry, allowing state-linked actors to rapidly produce highly convincing, localized content.
By publishing these findings, OpenAI aims to establish a precedent for transparency and collaboration in combating AI-driven disinformation. The company stated its goal is to “help our industry, governments, civil society and the public better identify and disrupt attempts by foreign threat actors to manipulate legitimate public debates, weaken democratic institutions and advance totalitarianism with AI characteristics.”
The disruption of these campaigns underscores the dual-use nature of advanced AI models. As technology becomes more powerful and accessible, the responsibility increasingly falls on developers like OpenAI to implement robust safeguards and threat intelligence capabilities to prevent their tools from being weaponized in the escalating US-China tech war.