- OpenAI banned two China-linked ChatGPT account clusters for spreading narratives against US AI infrastructure and tariffs.
- A Bitcoin Policy Institute report details China’s multi-vector influence operations, including over $2 billion routed to US advocacy groups.
- US officials and AI developers accuse China of industrial-scale AI model theft, leading Senator Tom Cotton to request a DOJ investigation.
OpenAI disclosed on June 10, 2026 that it had removed two clusters of ChatGPT accounts linked to China, used to run covert influence campaigns targeting US AI policy debates. The same day, Senator Tom Cotton wrote to Acting Attorney General Todd Blanche requesting a Department of Justice investigation into foreign efforts to shape American opinion on AI infrastructure. Taken together, the disclosures sketch a picture of a coordinated effort to slow US AI development through public narrative rather than direct technical competition.
PRC’s Multi-Vector Influence Targets US AI
A May 18, 2026 report by the Bitcoin Policy Institute identifies three converging vectors in what it describes as a multi-year foreign influence campaign against the US AI buildout: Chinese state media outlets, a network of US 501(c)(3) organisations aligned with the Chinese Communist Party, and foreign-tied charitable vehicles reportedly routing over $2 billion into US advocacy groups. The report argues these operations are designed to exploit and amplify existing domestic concerns, particularly around energy consumption and local opposition to data centers.
According to the report, these influence streams converged in support of the Sanders-Ocasio-Cortez AI Data Center Moratorium Act and various local moratoriums across the US. The pattern, as the report frames it, is straightforward: stoke opposition to US infrastructure while Beijing subsidises its own AI development at home.
Targeting AI Data Centers and Critical Infrastructure
OpenAI’s “Data Center Bandwagon” campaign involved generating social media content and images falsely claiming that AI data center construction was causing a “major spike in energy prices” for American families. The narrative tracks closely with coverage from Beijing’s English-language outlets. A March 2026 China Daily article, for example, described surging data center power demand as “a new source of economic strain” for US consumers.
The report also highlights foreign interference in the campaign against American AI.
AI Model Theft and Undermining US Innovation
The White House has accused foreign entities “principally based in China” of conducting “deliberate, industrial-scale campaigns” to extract capabilities from leading US AI systems through model distillation, a technique where a less capable model is trained on the outputs of a more powerful one. Major US AI developers have made parallel allegations. In February 2026, Anthropic accused DeepSeek and two other China-based AI laboratories of illicitly extracting capabilities from its Claude models through distillation. OpenAI sent a letter to US lawmakers stating it had found evidence of “ongoing attempts by DeepSeek to distill frontier models of OpenAI and other US frontier labs.”
Beyond the competitive implications, lawmakers and researchers have flagged a second concern: distilled copies of safety-tested models may not carry over the safety constraints built into the originals, potentially producing capable systems stripped of their guardrails.
US Lawmakers Demand Action and Transparency
In early June 2026, House Energy and Commerce Committee Chairman Brett Guthrie and other Republican lawmakers warned the Trump administration about potential CCP-backed entities supporting opposition to data center construction, calling for a federal investigation.
The central legal mechanism both letters point to is the Foreign Agents Registration Act. FARA requires individuals and organisations acting on behalf of foreign governments or principals to disclose that relationship publicly. Lawmakers argue that stronger FARA enforcement, combined with greater transparency in the funding of US advocacy groups, would help separate legitimate domestic opposition to AI development from campaigns coordinated abroad. The broader challenge of regulating AI activity across borders is one that US and EU policymakers are increasingly confronting from different angles.
Safeguarding American AI Leadership
The White House’s July 2025 AI Action Plan stated that the nation with the largest AI presence will “set global AI standards and reap the broad economic and military benefits.” That framing captures why influence over the public narrative around AI matters: if domestic opposition slows US data center construction or AI investment, the effect on competitive position is real regardless of whether that opposition was organic or amplified from outside.
Public concerns about AI’s effects on energy consumption, jobs and privacy are legitimate and deserve serious policy consideration. The difficulty for lawmakers is distinguishing those concerns from coordinated foreign messaging designed to exploit them. OpenAI’s account removals and the congressional calls for DOJ investigation suggest that distinction is now being treated as a national security question, not just a content moderation one. Whether FARA enforcement proves adequate to the scale of the problem remains to be seen. For more coverage of AI policy and regulation, visit our AI Policy & Regulation section.
