Iranian agents “spread false news in American media.” What is the story?

(CNN) - OpenAI announced Thursday that agents in Iran used its models and fictitious characters to publish news articles criticizing the US war on Iran in several US media outlets.

OpenAI explained in a post on its blog that about 12 media outlets around the world published nearly 100 articles written by 7 fake journalists.

The company said it blocked ChatGPT accounts that were directing the Persian-language AI tool to edit articles and editorial ideas, and generate comments on social media.

Some of the articles published with the help of artificial intelligence carried provocative titles, including: “Is American policy towards Iran privatized? The Jared Kushner question,” and “The war on Iran and the collapse of Western moral credibility.”

This is the latest example of the pivotal role played by artificial intelligence programs in the propaganda battles that coincide with the US-Israeli war on Iran.

Iranian youth used artificial intelligence to produce videos mocking US leaders, inspired by Lego bricks, while US President Donald Trump posted AI-generated videos depicting US strikes on Iran.

None of the publications that published these articles appear to be categorized as mainstream media, but some of them have a wide following online.

One of these outlets, the London-based Middle East Monitor, has nearly two million followers on Facebook.

Delivering anti-war rhetoric to the American public may be considered a successful media operation for those behind it.

OpenAI said it was unclear who was behind it, and whether it was affiliated with the Iranian government, and the company wrote in its blog: “The overall activity appears consistent with a commercial entity running a paid influence campaign, but we were unable to specifically identify the entity involved.”

The article about Kushner and US foreign policy was published in L.A. Progressive, which describes itself as an “independent, justice-driven publication” based in Los Angeles.

CNN has requested comment from LA Progressive.

Dawoud Abdullah, director of Middle East Monitor, told CNN that he would conduct an investigation into the articles spotted by OpenAI and take "appropriate" action.

According to OpenAI, after the articles were published, Iranian agents asked ChatGPT to generate comments that they could post on social media, claiming to be reactions to the articles.

“Users requested feedback on topics such as missile strikes, ceasefires, and diplomatic maneuvers, identifying deliverables that portray the United States as the aggressor, Israel as an unreliable ally, and Iran as a defiant but resilient victim,” she added.

“It is striking that the Iranians were able to publish their content in real media outlets, expanding the scope of their potential campaign,” said Jessica Brandt, who ran the Office of the Director of National Intelligence’s Center for Malign Foreign Influence (which was later closed) from 2023 to 2025.

Brandt added to CNN: “Today we are witnessing how artificial intelligence enables foreign adversaries to better hide their intentions and build more sophisticated false facades than they could otherwise do, in order to pass their content through trusted channels, exploiting these channels to reach a wider audience.”

The OpenAI findings show how artificial intelligence is increasingly being used to shape perceptions of war.

This spring, a false US intelligence report, powered by artificial intelligence, about an alleged Chinese ship transporting nuclear components, almost prompted the US military to board it, sources previously revealed to CNN.