Iran and China Conduct First Autonomous AI-Powered Influence Campaigns, Report Says
Cybersecurity researchers have documented what they describe as the first fully autonomous artificial intelligence influence campaigns carried out by state actors. According to analysis from security firms and intelligence researchers, operations attributed to actors in Iran and China each deployed AI systems that independently performed core tasks such as generating content, selecting targets, and adapting messaging based on audience response — all with significantly reduced human involvement compared to previous operations.
The campaigns mark a notable escalation from earlier AI-assisted influence efforts, which typically relied on humans to direct AI tools. In these new cases, researchers say the AI systems appear to have operated with a high degree of autonomy in crafting and distributing narratives across social media platforms and other online spaces.
The Iranian campaign was linked to inauthentic accounts that used AI to produce multi-language content targeting audiences in the Middle East, Europe, and North America. The Chinese operation similarly employed AI to generate content and manage personas across multiple platforms, researchers said.
Security analysts warn that autonomous influence operations could be harder to detect and more difficult to counter, as they can scale rapidly and adapt in near real-time without human review cycles. The development has prompted renewed calls among researchers and policymakers for improved detection methods and international norms around AI use in information operations.