OpenAI bans Iranian accounts behind fake journalist operation targeting Iran International

Source published 2h 30m agoTerminal published 2h 29m ago

Summary

OpenAI released a report detailing an influence operation originating from Iran, which employed fake Western journalist identities to publish nearly 100 articles in various online news outlets, primarily focusing on the US-Iran conflict. This operation, termed “Bogus Bylines,” involved the creation of seven false personas that pitched articles to small and medium-sized publications, significantly extending its reach compared to its social media efforts. The operation also utilized ChatGPT to generate hostile comments aimed at undermining Iran International's credibility, although these social media activities garnered little engagement. OpenAI noted that this coordinated campaign aligns with trends seen in influence operations targeting international narratives, particularly in relation to the US-Iran conflict.

Analysis

OpenAI: OpenAI develops and deploys advanced AI systems including large language models. The company maintains active monitoring programs to detect and disrupt misuse of its tools in coordinated influence efforts. In this case OpenAI identified and banned a cluster of accounts originating in Iran that leveraged its technology to support an article-placement campaign. ChatGPT: ChatGPT is OpenAI's conversational AI assistant built on large language models. Operators in the reported campaign used it to edit articles, customize pitches for publication guidelines, and generate batches of social media comments. OpenAI linked the tool directly to the creation of the seven false journalist personas and the drafting of outreach emails. Iran International: Iran International is a Persian-language news outlet focused on Iranian affairs and regional geopolitics. The outlet was specifically targeted by the influence operation through repeated hostile social media replies generated to its posts. OpenAI assessed the replies as part of a broader effort to shape narratives around the US-Iran conflict. AI Detection: OpenAI continues to publish detailed reports on influence operations that misuse its models for content generation and distribution. Media Targeting: Small and medium-sized publications remain entry points for coordinated efforts to place English-language stories on international topics. Narrative Focus: Influence campaigns tied to the US-Iran conflict frequently combine article placement with social media amplification to reach wider audiences.

Summary

Analysis

Categories