According to a Bloomberg report from October 14, 2026, networks linked to Russia, Iran, and China are utilizing generative AI to disseminate disinformation to American voters ahead of the November 3, 2026, midterm elections. These groups are leveraging mainstream tools, including OpenAI’s ChatGPT and Anthropic’s Claude, to influence public perception.
Russian-linked networks, such as Matryoshka, have been distributing AI-generated deepfakes of celebrities and fabricated news clips since early September 2026. Meanwhile, Iranian-associated groups are using AI to generate anti-Republican messaging and impersonate US activists. Chinese operations, including Spamouflage and Green Cicada, manage thousands of accounts to amplify narratives.
On October 8, 2026, OpenAI banned accounts associated with Russian and Iranian influence operations that utilized fake journalists and organizations. Despite these enforcement actions by companies like OpenAI and Meta, investigators note that detecting and removing content after it has been published remains a challenge, and the full scale of these campaigns is still under assessment.
Current intelligence assessments indicate that these operations aim to increase political polarization and undermine trust in the election process rather than directly targeting voting infrastructure. While Washington faces difficulties in implementing comprehensive countermeasures, the primary threat remains the manipulation of voter beliefs through AI-enhanced content.
