OpenAI Disrupts 5 AI-Powered, State-Backed Influence Ops

OpenAI identified five AI-enabled influence operations (Spamouflage, Bad Grammar, Doppelganger, IUVM, and Zero Zeno/Stoic) that used generative AI to produce and post political text across platforms like X, Telegram, Facebook, Instagram, Medium, and Blogspot; these campaigns were geographically diverse but low-impact (Brookings breakout score ≤2), and the report highlights difficulties in reliably detecting AI-generated text and the limits of technical defenses against coordinated disinformation.