The accounts in question reportedly used ChatGPT to produce long articles and short social commentary about both candidates to share on social media platforms.
But despite being posted on various websites, the influence operation “does not appear to have generated meaningful audience engagement,” according to a published warning.
OpenAI attributed the fake posts to a Tehran-backed group called Storm-2035, which and other Iranian groups have been raising alarms about over the past week as they continue to interfere in the election.
Google’s threat hunters have released intelligence on Iranian cyber influence activities following a recent spike in attacks that led to the leak of data from Donald Trump’s re-election campaign.
The artificial intelligence organization announced that it detected a total of 12 accounts on X and one on Instagram involved in this covert influence operation. In addition to the US presidential election, the fake news and commentary also covered the conflict in Gaza, Israel’s participation in the Olympics and other political issues.
“The first workflow produced articles about U.S. politics and global events that were published on five websites that acted as both progressive and conservative news organizations. The second workflow generated short commentaries in English and Spanish that were published on social media,” OpenAI said in a Friday statement.
The five domains are: niothinker.com, savannahtime.com, evenpolitics.com, theorator.com and westlandsun.com
Storm-2035 and one of its fake news sites, EvenPolitics, have been active online since the 2022 US midterm elections, publishing about 10 “articles” per week. These are written in English, but the group also has other influence-operations websites in Arabic, English, French and Spanish.
