OpenAI Exposes Chinese Influence Campaign Targeting U.S. AI Policy

Chinese Actors Manipulated AI Policy Discussions

OpenAI disclosed last week that actors linked to the People's Republic of China violated its terms of service to conduct covert influence operations aimed at shaping American technology policy debates. The company's report, "PRC-linked influence operations are targeting AI debates in the US," revealed that Chinese users exploited OpenAI's models to generate social media content designed to inflame controversies surrounding data center construction.

According to the findings first reported by The National Interest, one cluster of Chinese accounts produced comments and images claiming that data center expansion for AI infrastructure was driving up electricity costs for ordinary American families. A separate group created content attacking U.S. tariffs as tools of technological domination, with prompts specifically instructing the models to exclude references to Chinese President Xi Jinping while focusing criticism on President Donald Trump.

The influence campaign targeted an already contentious domestic debate over data centers—the energy-intensive facilities essential to training and deploying advanced AI systems—that has generated friction among technology companies, local communities, environmental activists, and policymakers at every level of government.

Why it matters

This disclosure arrives as China demonstrates accelerating technical capabilities in artificial intelligence, transforming what was once a comfortable U.S. lead into a competitive race with national security implications. Foreign manipulation of domestic technology policy debates could slow American infrastructure development precisely when maintaining technological advantage matters most.

China's Technical Advances Narrow the Gap

The influence operation coincides with significant shifts in the global AI landscape. Stanford University's Institute for Human-Centered Artificial Intelligence reported in April that China has nearly eliminated the United States' longstanding advantage in AI development. The researchers noted that China's DeepSeek-R1 model achieved performance parity with America's most advanced systems in February 2025, while the gap between Anthropic's leading model and its Chinese equivalent has shrunk to just 2.7 percent—previously a double-digit margin.

Simultaneously, the House Select Committee on the Chinese Communist Party held hearings in April examining Chinese efforts to acquire American AI technology through illicit means. Committee Chairman John Moolenaar highlighted a $2.5 billion chip smuggling case that would constitute the largest export control violation in U.S. history.

Progressive Lawmakers' Proposals Align with Beijing's Goals

Some prominent progressive members of Congress have advanced legislative proposals that align with the narratives Chinese influence operations sought to amplify. Senator Bernie Sanders and Representative Alexandria Ocasio-Cortez introduced the "AI Data Center Moratorium Act" in March, arguing against allowing "a handful of billionaire Big Tech oligarchs" to make decisions reshaping the economy and democracy.

Sanders subsequently hosted a Capitol Hill discussion featuring academics from Tsinghua University and the Beijing Institute of AI Safety and Governance—a decision that drew criticism even from the Washington Post editorial board, which characterized the approach as "dangerous" given China's human rights record.

OpenAI emphasized in its disclosure that while the Chinese operation attempted to exploit legitimate public concerns about energy costs and local impacts of data center development, the company found no evidence the campaign achieved meaningful reach beyond its own coordinated activity.

The details were first reported by The National Interest in an analysis by Michael Rosen, a nonresident senior fellow at the American Enterprise Institute specializing in technology and intellectual property issues.