OpenAI Disrupts AI‑Powered Disinformation Campaigns Linked to Russia and Iran

OpenAI announced that its internal safety team has successfully intervened in two separate influence operations that exploited its language model, ChatGPT, to spread disinformation. The interventions targeted a Russian‑run propaganda effort aimed at Latin America and a network of Iranian actors that used the model to bolster fabricated journalist personas.

According to the company's statement, the Russian operation leveraged ChatGPT to generate and amplify political content across social media platforms, seeking to shape public opinion in several Latin American countries. The content, which included articles, posts, and comment threads, was designed to appear organic while subtly promoting narratives favorable to Russian strategic interests.

In a parallel effort, Iranian operatives employed the same AI tool to enhance a collection of fake journalist identities. By using ChatGPT to produce polished articles and commentary, the actors attempted to lend credibility to their misinformation campaigns, making it harder for readers to discern authentic reporting from state‑sponsored fabrications.

OpenAI said the disruptions were carried out after its safety team identified patterns of misuse that violated the company's usage policies. The team worked with platform partners and relevant authorities to curtail the accounts and content involved. While the company did not disclose the precise technical methods used to detect the abuse, it highlighted ongoing investments in monitoring tools designed to spot coordinated inauthentic behavior powered by large language models.

The revelations come amid growing concern among policymakers and technology firms about the role of generative AI in amplifying misinformation. Experts note that the ease of producing high‑quality text at scale lowers the barrier for state and non‑state actors to launch sophisticated influence campaigns. OpenAI’s actions illustrate a broader industry trend toward more proactive moderation and collaboration with external stakeholders to mitigate AI‑enabled threats.

OpenAI has indicated that it will continue to refine its safeguards and work closely with governments and civil‑society groups to address the misuse of its technology. The company’s effort to dismantle these operations underscores the challenges of balancing open AI development with the responsibility to prevent harmful applications, a debate that is likely to shape regulatory discussions in the months ahead.