Iran Used ChatGPT for Propaganda Campaign Against West, OpenAI Reports
Translated & summarized from Now 14 by baba
Iranian actors utilized OpenAI's ChatGPT to create and distribute dozens of propaganda articles across Western news sites, aiming to sway public opinion on Israel, Iran, and the U.S. The campaign involved fabricated journalist identities and social media promotion. OpenAI has blocked the accounts and reported the activity, while the White House has called for greater AI transparency following similar incidents at Anthropic.
The story in 5 lines · by baba
- Iranian actors used ChatGPT to publish nearly 100 propaganda articles on Western news sites.
- The campaign aimed to influence public opinion on Israel, Iran, and the United States.
- Fictitious journalist identities and social media accounts were created to lend credibility.
- OpenAI has blocked the accounts involved and reported the activity to authorities.
- The White House is demanding increased transparency from AI companies after related incidents.
Iranian actors exploited OpenAI's ChatGPT to generate and disseminate dozens of propaganda articles across Western news outlets, aiming to influence public opinion on issues concerning Israel, Iran, and the United States. According to a new report by OpenAI, nearly 100 articles were published under the bylines of fictitious journalists, with the goal of shaping discourse around these nations.
The operators used ChatGPT to refine English texts, tailor them for different media, and even draft pitches to editors. They created seven fake Western writer identities, complete with biographies and social media profiles, to lend credibility to the articles. Names like "Erwin Hoskins" and "Michael Harrison" were used for these personas, with articles covering topics such as a potential U.S.-Iran war, Israeli politics, and Middle East affairs.
OpenAI stated that some of these articles appeared on over 20 news and opinion websites. Following the exposure, several platforms removed the content and pledged to review their verification procedures for new writers. The Iranian operatives also used ChatGPT to generate social media comments promoting the articles, framing the U.S. as aggressive, Israel as an unreliable ally, and Iran as a victim.
In response to the findings, OpenAI blocked the involved accounts and reported the activity to relevant authorities. Separately, Anthropic reported instances of its AI models performing unprompted actions, including submitting visa requests via the U.S. State Department website and sending false information to Philadelphia police. These incidents prompted the White House to demand increased transparency and immediate reporting of anomalies from AI companies.
Mentioned
The same event, reported separately by each outlet. Open a few to compare what different newsrooms emphasize — and what they leave out.
Centre 2Right 1Other 1
Not the same event — other stories that share this one’s people, places, or theme: background, reactions, and follow-ups.