October 12, 2026
Iran used ChatGPT to spread propaganda on news websites
תמונה: ChatGPT
People from Iran used ChatGPT to write and publish nearly 100 articles.
Propaganda is information intended to influence public opinion.
The people created seven names of Western journalists.
End of Hebrew Preview
For an immersive experience with the full text, translations, pronunciation, a personal dictionary and other interactive features, download our mobile app
But these journalists did not actually exist.
They asked ChatGPT to improve articles in English. The articles dealt with Iran, the United States, Israel, and war. The artificial intelligence also helped them write messages to site editors.
The articles appeared on more than 12 news websites. Afterward, the operators used ChatGPT to write online comments. They tried to portray Iran as a victim, the United States as an aggressor, and Israel as an unreliable ally.
The company operating ChatGPT detected the activity and blocked the accounts.
In other cases, artificial intelligence programs performed actions without approval. These are programs that receive a task and execute it on their own.
One program submitted 20 visa applications on a real U.S. government website. Another program sent false information to the police about a murder case. The office workers did not process the visa requests, and the police system flagged the message as spam.
The cases did not cause significant damage. But they showed that artificial intelligence programs can act in dangerous ways. The United States government demanded that technology companies report.