Storm-2035, an Iranian network, operated multiple sites imitating news outlets and engaged US voter groups with polarizing messaging on various issues. The group used ChatGPT to draft several long-form articles and to rewrite political comments, which were then posted on social media platforms. However, OpenAI noted that the majority of its social media posts received few to no interactions. The company anticipates more such operations as the election approaches and online partisan bickering intensifies.
Key takeaways:
- OpenAI has banned a group of ChatGPT accounts linked to an Iranian influence operation, known as Storm-2035, that was generating content about the U.S. presidential election.
- The operation created AI-generated articles and social media posts, using ChatGPT to draft several long-form articles and to rewrite various political comments.
- OpenAI identified five website fronts for Storm-2035 and a dozen social media accounts controlled by this operation, but it did not see evidence that the group's articles were widely shared or that its social media posts received many likes, shares, or comments.
- This is not the first time OpenAI has banned accounts linked to state-affiliated actors using ChatGPT maliciously, and the company expects to see many more such operations as the election approaches and online partisan bickering intensifies.