OpenAI disrupts two influence operations linked to Russia and Iran

14:30

By: Azzat Manal

OpenAI disrupts two influence operations linked to Russia and Iran

OpenAI says it has identified and disrupted two influence operations that allegedly used ChatGPT to produce and distribute misleading content about geopolitical conflicts. The campaigns, known as “Dark Clark” and “Bogus Bylines,” were linked respectively to Russian and Iranian actors, according to the company’s latest report on the misuse of artificial intelligence.

The findings highlight growing concerns about how generative AI tools can be incorporated into coordinated information campaigns. While such systems can assist with writing, translation and research, they can also be misused to prepare deceptive narratives, fabricate identities and adapt content for different audiences.

According to OpenAI, the Dark Clark operation targeted audiences in several Latin American countries with content intended to undermine Ukraine and criticise certain regional political leaders. Among the activities identified was the circulation on TikTok of an allegedly fabricated audio recording attributed to Ukraine’s consul in Ecuador. The operators also reportedly used a fake email address to encourage schools in Peru to organise events connected to Ukraine.

OpenAI said it assessed the campaign as level five on its six-point scale for measuring the impact of influence operations. The company nevertheless cautioned that some claims made by the individuals involved could not be independently verified, underlining the difficulty of determining the full reach and effectiveness of covert online activity.

The second operation, dubbed Bogus Bylines, allegedly focused on producing misleading articles about the conflict between the United States and Iran. OpenAI said ChatGPT was used to refine drafts, adapt articles to the editorial requirements of online publications and prepare submission emails intended to secure publication.

The campaign reportedly relied on seven fictitious authors who presented themselves as Western journalists. OpenAI said it had identified nearly 100 articles published across more than a dozen media outlets since July 2025. The reported activity illustrates how fabricated identities and AI-assisted writing can be combined in an attempt to make misleading material appear credible.

The use of established media platforms can give deceptive content access to audiences beyond those reached through newly created social media accounts. Articles published on recognised websites may also be shared elsewhere, quoted by other users or incorporated into wider online discussions, potentially making the original source and purpose of the content more difficult to identify.

However, the publication of an article does not by itself establish how many people encountered it, whether readers believed its claims or whether it changed public opinion. Assessing the actual influence of such campaigns requires evidence about distribution, audience engagement and the wider information environment.

The cases also raise questions about the role of artificial intelligence companies in identifying and limiting coordinated abuse. OpenAI says it monitors suspicious activity and takes action against accounts that violate its policies. Such measures form part of a broader effort across the technology sector to detect automated manipulation, impersonation and the deliberate spread of false information.

At the same time, the growing availability of AI writing tools complicates efforts to distinguish legitimate political commentary from organised deception. Investigators and platforms must consider not only the content itself but also patterns of behaviour, coordination between accounts, fabricated identities and attempts to conceal who is behind a message.

OpenAI’s findings point to an evolving form of digital propaganda in which AI can support the production and adaptation of misleading material, while existing social networks and online publications provide potential channels for distribution. The report underscores the importance of transparent sourcing, independent verification and stronger cooperation among technology companies, media organisations and researchers investigating influence campaigns.