Iranian campaign uses ChatGPT to plant fake articles in US news outlets

An Iranian-linked influence campaign used OpenAI’s ChatGPT to produce and place approximately 100 articles across at least 20 news publications, raising concerns about the use of artificial intelligence to influence public opinion through established media platforms.

According to a report published by The Washington Post on Friday, October 9, 2026, OpenAI identified the operation, which involved AI-generated articles appearing in legitimate news outlets.

The articles reportedly included commentary critical of the United States’ war with Iran, with some appearing under the same byline across multiple publications.

One example was an opinion article published on September 22 by the Port St. Joe Star, a small weekly newspaper in Florida.

The article, titled “The Cost of American Power in the Iran War,” appeared under the name Ervin Hoskins and warned about the financial consequences of the conflict for Americans.

The same byline reportedly appeared on other articles critical of the war in publications including Daily Kos and Middle East Monitor, which collectively have more than three million Facebook followers.

In its October 8 report, OpenAI identified the campaign as “Bogus Bylines,” saying its operators used seven fictitious journalist identities to submit articles to publications.

The company said the operators used ChatGPT to edit articles, prepare pitches to editors and generate social media comments. Some accounts supporting the fabricated identities were also created on platforms including Facebook, Instagram and X.

OpenAI traced the activity to accounts originating in Iran but said it could not establish the identity of the organisation responsible or confirm Iranian government involvement.

The company subsequently banned the accounts and shared information about the operation with relevant authorities.

The findings have raised concerns about the ability of AI-assisted influence campaigns to exploit legitimate publishing channels, particularly when articles appear under apparently credible journalistic identities.

The case also highlights the challenges facing news organisations in verifying contributors and identifying coordinated attempts to influence public debate.