Private companies in Iran, China, and Israel are accused of using AI agents to run social media influence operations

U.S. officials said private companies in Iran, China, and Israel ran some of the first social media influence operations executed autonomously by AI agents. According to The New York Times, these operations combined downloadable Chinese open-source models with agent software that required no human input, creating fake accounts in bulk on Instagram, Facebook, X, and TikTok and posting fabricated content targeting political issues and current affairs; operators also removed built-in model restrictions meant to prevent generating fake identities or distorting online discussions. Investigators determined the Chinese and Iranian operations had government support, while the two Israeli operations were linked to private companies rather than the state. Separately, The Guardian reported that a pro-Israel website run under the name of a nonexistent think tank published 124 reports totaling over 560,000 words in nine days; its registrant, Piro Inc., has registered with the Department of Justice under the Foreign Agents Registration Act as distributing content on behalf of the Israeli government, and its underlying technology is a commercial service marketed around generating content easily quotable by AI chatbots.