OpenAI uncovers PRC-linked influence campaigns targeting AI policy debates

OpenAI documents state-backed influence operations

OpenAI has published an intelligence report exposing coordinated influence operations linked to actors from the People’s Republic of China (PRC). The report documents campaigns that used AI tools as an operational instrument for shaping public debate about artificial intelligence policy in the United States.

This document represents an unusual public window into how a company developing some of the world’s most widely used AI systems actively monitors and disrupts attempts to misuse its own technology for disinformation purposes.

What did the influence campaigns target?

According to the report’s findings, the campaigns focused on several key topics of strategic interest to the PRC in the context of global technology competition:

• AI policy in the US — efforts to shape public opinion and political debate on the regulation and development of artificial intelligence
• Data center narratives — the broader geopolitical and economic implications of the infrastructure essential for AI
• Tariff policy — in the context of ongoing trade tensions between the US and China
• False claims about ChatGPT — disinformation aimed directly at the reputation and trustworthiness of OpenAI’s leading product

How generative AI is used as an operational weapon

The report’s key finding is not merely that disinformation is being spread about AI — but that generative AI has itself become an operational tool within information warfare. The campaigns did not use AI solely as a subject, but as a means of scaling operations: accelerating content production, personalizing messages, and distributing them at volume.

A new view of AI threats

Until now the prevailing narrative was that AI systems could be a source of disinformation through hallucinations or through misuse by anonymous actors. This report introduces a different dimension: organized state-linked actors are actively integrating generative AI into operations that target the emerging techno-political space.

The development and regulation of AI technology is not just a question of innovation and technical safety — it has become an arena of geopolitical competition in which state actors with a strategic interest in who develops frontier AI are fighting for narrative dominance.

OpenAI as a security actor

The publication of this report positions OpenAI not merely as an AI company, but as an active security policy actor that monitors threats, conducts internal investigations, and publicly releases findings. This is in line with a trend we also see at other major technology companies — Microsoft, Google, and Meta regularly publish so-called threat intelligence reports on state-sponsored threats.

Unlike traditional cyberattacks that target infrastructure, these operations target something more subtle: public discourse and policy-making. Precisely because they are harder to detect and leave fewer clear technical traces, such operations represent a particular challenge for defensive mechanisms.

Broader context: AI policy as a geopolitical battlefield

This report comes at a moment when AI regulation in the US is the subject of intense legislative activity, and the relationship between the US and the PRC in the field of high technology is one of the defining geostrategic questions of our era. When these two variables converge — rapid development of an AI regulatory framework and active geopolitical competition — it is natural that the information space around AI policy becomes a target.

OpenAI does not reveal the full scope of detected operations, but the very act of making this public sends a clear message: companies developing AI systems also have a responsibility to monitor and suppress their misuse, including misuse by state actors who use those same systems as an operational weapon.