OpenAI report: PRC-linked operations used ChatGPT to stir U.S. AI data center and tariff debates

OpenAI published a threat intelligence report on June 10, 2026 documenting two separate influence operations linked to Chinese state-affiliated actors that used ChatGPT to generate content targeting U.S. policy debates on AI infrastructure and trade. The campaigns — both disrupted by OpenAI — are the first the company has publicly documented that targeted the AI data center debate specifically.

What's new

OpenAI identified and terminated accounts associated with two distinct operations:

• "Data Center Bandwagon": Accounts believed to be linked to a Chinese government contractor used ChatGPT to generate social media posts and comic strips claiming AI data centers were driving up electricity prices for American families. Content focused on U.S. power grid capacity and framed AI infrastructure growth as a burden on ordinary consumers.
• "Tech and Tariffs": A separate operation used ChatGPT to create political cartoons and written content criticizing Trump administration tariffs and the U.S. push for global technology dominance.

Neither campaign gained significant traction on social platforms before being disrupted. Ben Nimmo, principal investigator on OpenAI's intelligence and investigations team, described the activity as an influence operation trying to interfere in an already-existing debate rather than creating one.

The report also documents false claims circulated about ChatGPT itself, as part of a broader effort to undermine trust in U.S. AI platforms.

Context

OpenAI has published influence operation disruption reports periodically since 2024, covering state-linked campaigns across multiple countries. Prior reports documented operations targeting electoral politics and international news narratives. This marks the first time OpenAI has publicly attributed a China-linked influence campaign to the domestic U.S. debate over AI data center energy use and infrastructure policy.

The timing matters: U.S. AI data center expansion has become a genuine flashpoint for debates over energy prices, grid reliability, and federal permitting. The AI industry's push for infrastructure exemptions and the massive power demands of frontier model training have generated real public controversy — creating fertile ground for influence operations that amplify existing tensions.

Why it matters

This report confirms that AI policy debates themselves — not just elections or foreign policy — are now targets of state-linked influence operations.

AI companies as both vectors and targets: ChatGPT was used to produce the influence content, making OpenAI simultaneously the subject of false claims and the platform that generated the disinformation. This dual role puts unique pressure on trust and safety infrastructure.

Detection is harder: These operations exploit already-existing debates, meaning content often closely mirrors legitimate domestic opposition. Separating synthetic influence from organic political speech requires attribution inference, not just content analysis.

Policy implications: As Congress and federal agencies consider legislation around AI data centers, energy policy, and national security, AI-generated opposition content complicates the legislative signal. Lawmakers may now face foreign-sourced content that appears to be genuine domestic sentiment.

OpenAI's willingness to publish this report — naming PRC-linked actors specifically — reflects a broader industry trend toward public attribution, following similar disclosures by Google, Meta, and Microsoft.