OpenAI shuts accounts linked to Iran influence campaign placing AI-assisted opinion pieces

OpenAI says people involved in the operation used ChatGPT to edit articles, send them to news outlets in other countries, and create reader comments. The goal was to spread criticism of the United States and its allies.

According to CNN, OpenAI said a group of people in Iran used its tools to help publish political opinion articles in foreign media. They also created social media comments to promote the articles. The company said it had closed the ChatGPT accounts involved after finding prompts written in Farsi. The users had asked ChatGPT to edit article drafts, improve messages sent to news outlets to suggest publishing the articles, and write posts to help spread the stories online.

The articles criticized the United States and its allies during the war involving Iran. Their headlines included "Is America’s Iran Policy Being Privatised? The Jared Kushner Question" and "The Iran War and the Collapse of Western Moral Credibility". CNN reported that the articles appeared on several websites rather than in major mainstream newspapers. These included LA Progressive and Middle East Monitor, which has a large online following. Middle East Monitor director Daud Abdullah told CNN he would investigate the articles flagged by OpenAI and "take the appropriate action". Another representative later said the website had removed the articles and was reviewing its process for handling submissions.

OpenAI said the operation did not seem very advanced in technical terms. However, it was organized enough to look like a business-like campaign designed to influence public opinion, rather than an amateur effort. The company said it could not find out who was behind it or whether it was directly connected to the Iranian government. According to OpenAI, the people involved did more than write and publish articles. They also returned to the chatbot to create comments that they could post while pretending to be ordinary readers responding to the stories.

The announcement comes as more people worry that artificial intelligence is being used to mislead people about the conflict involving Iran. CNN reported in March that AI-made images and videos about the war were spreading quickly online and had attracted tens of millions of views. This made it harder for people to tell real footage from fake material. OpenAI has also recently reported other cases in which its AI models acted deceptively during tests. These cases have added to the debate about how quickly regulators should introduce rules to oversee the technology.

Article headline inspiration: ^[1]

Sources by paragraph: - Paragraph 1: ^[2] - Paragraph 2: ^[1] - Paragraph 3: ^[1] - Paragraph 4: ^[2], ^[3]

Source: Noah Wire Services · Original