A covert influence operation, meaning a coordinated effort to shape what people believe while hiding who is behind it, was found using OpenAI's ChatGPT. OpenAI identified a set of Russian accounts running the campaign and shut them down.
The short version: Russian accounts were using ChatGPT to support a disinformation effort, and OpenAI removed their access.
What the company actually said
OpenAI described the episode in specific terms, saying the incident "illustrates how influence actors can use AI as a supporting tool within a broader effort to manufacture authority."
Influence actors, in that framing, are groups running coordinated, often hidden campaigns to shift what people believe. Manufacturing authority means creating the look of credibility rather than earning it, making content appear expert or established when it is not.
OpenAI's account separates the campaign's human drivers from the AI tool they used. The people behind the operation ran it. ChatGPT was one resource they reached for.
What OpenAI confirmed and what it left open
OpenAI confirmed it identified the Russian accounts and banned them. The company did not say how many accounts were involved, how long the campaign ran, or what content it generated.
Publishing a public statement rather than handling this as a routine enforcement action signals that OpenAI sees the episode as worth naming openly. In the company's framing, influence operations are now incorporating AI as a supporting instrument, one that helps produce material designed to look authoritative.
The accounts are gone. OpenAI's statement describes this as an example of influence actors using AI to manufacture authority.