OpenAI Bans Accounts Tied to New PRC Influence Operation

OpenAI has banned accounts linked to a previously unreported influence operation originating from the People's Republic of China, a campaign the company dubbed "Nine-emdash Line."

The operation used AI tools to generate regional influence content focused on three politically charged subject areas: the South China Sea, Hong Kong, and US politics. OpenAI disclosed the takedown in its reporting on threat activity, marking the first public identification of this specific network.

The name references the "nine-dash line," Beijing's cartographic claim to most of the South China Sea — the maritime region at the center of territorial disputes between China, the Philippines, Vietnam, and other neighboring states. The name signals the operation's core thematic focus on advancing narratives favorable to Beijing's position in those disputes.

The campaign's scope spanned both foreign and domestic political targets. Content on the South China Sea and Hong Kong addresses issues where Beijing has direct geopolitical stakes, including sovereignty claims and its governance of the former British colony. The inclusion of US politics in the operation's content mix extends its reach into Western democratic discourse, a pattern consistent with previously documented PRC-linked influence networks.

The disclosure matters beyond a single takedown. AI-generated content lowers the cost of producing persuasive, localized material at scale, and OpenAI's platform has become a recurring instrument for state-linked influence actors. Each ban like this one offers researchers and policymakers another data point on how generative AI is reshaping state-directed information operations — the subject of intensifying scrutiny from US lawmakers and allied governments.

OpenAI did not state how many accounts it banned, how long the operation ran before detection, or which specific narratives the content promoted. The company has previously reported disrupting similar PRC-linked networks that used its models for influence activity, alongside operations attributed to Russia, Iran, and other state actors.

Expect continued disclosures of this kind: OpenAI has committed to publishing regular updates on influence operations it detects and disrupts, making its threat reports a running record of how state actors adapt AI tools to information warfare.