Iranian operatives used ChatGPT to publish anti-US war articles: OpenAI
What's the story
Iranian operatives have allegedly exploited artificial intelligence (AI) tools to publish anti-US war articles in several media outlets, OpenAI has revealed. The operation involved seven fake journalists and around 100 articles across a dozen global publications. OpenAI has since banned the ChatGPT accounts involved in this campaign, which included prompting the AI tool in Persian for article revisions, editorial pitches, and social media comments.
Propaganda strategy
Articles placed in real publications
The use of AI tools in this operation highlights a new method of propaganda dissemination.
Not only were these tools used to create misleading content, but they also helped place it in real publications, giving false narratives a wider reach.
Some articles featured provocative titles like "Is America's Iran Policy Being Privatized? The Jared Kushner Question" and "The Iran War and the Collapse of Western Moral Credibility."
Media reaction
Operation consistent with commercial actor conducting influence campaign for hire
OpenAI couldn't determine who was behind the operation or if they were affiliated with the Iranian government.
The activity appeared consistent with a commercial actor conducting an influence campaign for hire.
After the articles were published, users asked ChatGPT to generate comments that could be posted on social media as if they came from people reacting to the coverage.
The requested comments addressed subjects including missile strikes, ceasefires, and diplomatic maneuvers.
AI implications
AI could help foreign actors hide their involvement
Jessica Brandt, former head of the US Office of the Director of National Intelligence's Foreign Malign Influence Center, said AI could help foreign actors hide their involvement and facilitate the spread of their messaging through legitimate channels.
This case comes amid a wider struggle to shape public perceptions of the US and Israel's war with Iran.