AI-generated deepfake videos were used to create fake New Yorkers criticizing mayor-elect Zohran Mamdani through accounts called 'Citizens Against Mamdani' on multiple social media platforms, demonstrating automated astroturfing at scale.
Unknown actors created social media accounts called 'Citizens Against Mamdani' on TikTok, Twitter, and Instagram that used AI-generated deepfake videos to create fake New Yorkers criticizing mayor-elect Zohran Mamdani. The videos, which appear to use OpenAI's Sora technology based on visible watermarks, feature diverse fake personas with New York accents making claims about Mamdani's alleged anti-Americanism, tax plans, and false promises on rent and transportation. The videos have accumulated tens of thousands of views, with the TikTok account gaining about 30,000 likes. Reality Defender analyzed several videos and found a 99% likelihood that at least one video was a deepfake. The campaign represents what experts call automated astroturfing, where generative AI democratizes the creation of false grassroots political opposition. Unlike previous astroturfing efforts that required hiring human influencers or operating bot networks, this campaign demonstrates how AI can generate realistic-looking people and scripts at scale with minimal human effort. The incident highlights how the cost of producing political disinformation has dramatically decreased, making such campaigns accessible to a wider range of actors.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Using AI systems to conduct large-scale disinformation campaigns, malicious surveillance, or targeted and sophisticated automated censorship and propaganda, with the aim of manipulating political processes, public opinion, and behavior.
AI system
Due to a decision or action made by an AI system
Intentional
Due to an expected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
