Brief
OpenAI says Iranian agents used ChatGPT to generate fake news articles
OpenAI says the operation created seven fake Western journalist profiles and posted about 100 AI‑generated articles between July 2025 and September 2025.
By Felo News Desk · Published
OpenAI announced on Thursday, October 8, that agents of the Iranian regime employed its ChatGPT tool to produce fake articles that criticised the United States’ war against Iran.
The company said investigators uncovered seven fabricated Western journalist personas that were used to pitch AI‑generated pieces to 12 small‑to‑medium news outlets. Almost 100 articles were published under those bylines from July 2025 through September 2025.
OpenAI blocked the ChatGPT accounts used for the operation and noted that the fake stories were accompanied by positive social‑media comments to simulate popularity. Sample headlines included “Trump’s Iran Gamble Hardened the Regime it was Meant to Break” and “Why Trump Blinked on Iran.”
Middle East Monitor’s editor‑in‑chief Dawoud Abdullah told CNN he would review the identified pieces and take “appropriate action.” A later outlet representative said the articles had been removed and the submission process was under review.
Key facts
- OpenAI identified seven fake journalist personas used by Iranian agents. (voanews.com)
- Nearly 100 AI‑generated articles were published in 12 outlets between July 2025 and September 2025. (voanews.com)
- OpenAI blocked the ChatGPT accounts involved and reported the operation to the media. (voanews.com)
- Middle East Monitor’s editor said the outlet would review and remove the articles. (voanews.com)
Sources
- [1] voanews.com — originally reported as “Open AI says Iranian regime used ChatGPT to create fake articles”
