October 9, 2026. OpenAI says it caught an Iranian-linked influence operation using ChatGPT and fake personas to plant roughly 100 articles, critical of the U.S. war with Iran, in media outlets around the world — about a dozen publications ran the AI-assisted pieces.
What happened
On Thursday, OpenAI published findings on an influence campaign it disrupted: operatives in Iran used ChatGPT accounts and seven fabricated journalist personas to get articles into real outlets. The company said it banned the ChatGPT accounts after discovering they were prompted in Farsi to draft and tweak articles and editorial pitches, and to generate social media comments amplifying them.
Some pieces ran under provocative headlines, including “Is America’s Iran Policy Being Privatized? The Jared Kushner Question” and “The Iran War and the Collapse of Western Moral Credibility.” The named outlets included the London-based Middle East Monitor — nearly two million Facebook followers — and the LA Progressive in Los Angeles.
The details that matter
After the articles were published, the operatives asked ChatGPT to produce comments they could post on social media pretending to be readers reacting to the stories — on topics like missile strikes and ceasefires — framed to portray the U.S. as the aggressor, Israel as an unreliable ally, and Iran as a resilient victim. OpenAI said it could not confirm whether the Iranian government was behind the campaign, noting the activity was “consistent with a commercial actor running a for-hire influence campaign.” None of the publications that ran the articles appear to be major mainstream outlets, though several have sizable audiences.
Jessica Brandt, who ran the Office of the Director of National Intelligence’s Foreign Malign Influence Center from 2023 to 2025, told CNN it was “remarkable” the operatives landed content in real media outlets — and that AI is letting foreign adversaries launder content through authentic channels to hide their hand.
Why it matters
This is the shape of modern influence operations: not bot swarms screaming into the void, but a small team using AI to write, pitch, and amplify material polished enough for real publications to run. The fact that OpenAI — the company whose models powered the campaign — caught and disclosed it is both reassuring and unsettling: detection works, but the campaign worked first. For anyone reading AI-era news, the takeaway is blunt: authorship is now the easiest thing in the world to fake.
FAQ
What did OpenAI say happened?
Iran-linked operatives used ChatGPT and seven fake journalist personas to place about 100 AI-assisted articles critical of the U.S. war with Iran in roughly a dozen outlets worldwide. OpenAI banned the accounts.
Was the Iranian government behind it?
OpenAI said it could not confirm that. The activity was consistent with a commercial actor running a for-hire influence campaign.
Why does this matter beyond Iran?
It shows AI is making foreign influence campaigns cheaper and harder to spot — fake personas and AI-assisted pitches got real publications to carry the content.
Sources: CNN, OpenAI blog post