Iran and China Deploy Autonomous AI Influence Campaigns

The governments of Iran and China, alongside private firms in Israel, have deployed the first autonomous AI influence campaigns on platforms including X, TikTok, Facebook, and Instagram. These actors utilized open-source Chinese AI models to create agents capable of independently opening networks of fake accounts and seeding them with political content with minimal human intervention.

Iran launched a state-backed campaign targeting U.S. audiences, where AI agents posed as American citizens to spread anti-Republican views. Simultaneously, the Israeli firm IntelEye used the Chinese model DeepSeek to control thousands of accounts discussing Israeli national elections. Co-founder Maor Sellek stated the activity was defensive research intended to test AI safety and improve detection of such activity.

Meta detected and removed the inauthentic accounts, describing the activity as a coordinated attempt to create fake profiles across several platforms. U.S. officials and security researchers warned that these agentic campaigns bypass traditional AI safeguards and could be used to target voters in the upcoming November midterm elections.