This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
Following the capture of Venezuelan leader Nicolás Maduro in January 2026, a surge of AI-generated images and videos circulated on social media platforms like X, TikTok, and Instagram. These synthetic visuals, which included fabricated scenes of Maduro in U.S. custody and crowds celebrating his removal, reached millions of viewers and were amplified by political figures. The incident highlights the increasing use of generative AI tools to manufacture consent and spread misinformation during high-stakes geopolitical events.
Following the January 2026 U.S. raid and arrest of Venezuelan leader Nicolás Maduro, purportedly AI-generated images and videos circulated widely on platforms including X, reportedly depicting Maduro in fabricated scenarios. NewsGuard identified at least seven manipulated or synthetic visuals reaching over 14 million views, including AI-generated photos detected via Google Gemini's SynthID watermark.
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.
Also classified as
The incident involved the generation of highly realistic false information that misled millions of social media users.
The incident involves the coordinated and at-scale dissemination of AI-generated deepfakes and images to manipulate public opinion and manufacture consent regarding a geopolitical conflict.
Human
Due to a decision or action made by humans
Intentional
Due to an expected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The risk was caused by human actors intentionally generating and spreading AI-generated misinformation post-deployment to manipulate public opinion.
Added categories peak at 4: Severe (Epistemic Harm, Direct)
Following the capture of Venezuelan leader Nicolas Maduro, a massive AI-generated disinformation campaign flooded social media, using hyperrealistic deepfakes to manufacture consent for foreign intervention and manipulate public opinion. Amplified by high-profile figures, the incident highlights the potent threat of generative AI as an information warfare tool during geopolitical crises.
The reports describe generative AI systems being used as tools by humans to quickly generate hyper-realistic video and images from text prompts, without the AI exhibiting independent planning or autonomous decision-making.
Limited Risk: The reports describe the generation of deepfakes and AI-generated content, which under the EU AI Act are subject to transparency and labeling obligations to ensure users are aware they are interacting with or viewing AI-generated media.
The AI systems were used to generate hyperrealistic fake videos of Venezuelan citizens celebrating and fake images of Maduro in custody to mislead the public.
The reports provide consistent, detailed accounts of the AI-generated misinformation campaign, including specific platforms (X, TikTok, Instagram), view counts (14 million), and the specific types of synthetic media generated. The role of AI tools in creating this content is explicitly stated and verified by fact-checking organizations like NewsGuard, BBC, and AFP.
None. The incident is clearly a case of politically motivated AI-generated disinformation.
Technology
Hyperrealistic Generative AI Tools
Tools like Sora and Midjourney make creating realistic fake videos easy.
Inadequate Automated AI Detection
Detection and labeling tools yielded mixed results across major platforms.
Imitation of Reality Improvements
AI is getting better at imitating reality, outpacing detection tech.
Data Inputs
Lack of Media Fingerprinting
Absence of real media verification makes tracking authentic content hard.
Use of Simple Text Prompts
Simple text prompts are used to generate hyperrealistic visuals easily.
Human Factors
Confirmation Bias of Users
Netizens shared AI content to manufacture support for political narratives.
High-Profile Amplification
Influential figures and politicians reshared unverified AI-generated slop.
Viewer Inability to Spot Fakes
Hyperrealistic visuals blur the line between fiction and reality for users.
Process and Methods
Slow Fact-Checking Mechanisms
Community Notes and fact-checkers react too slowly to stop viral spread.
Inconsistent Platform Labeling
Some platforms labeled AI videos while others missed fabricated content.
Management
Prioritizing Engagement Over Safety
Social media platforms allowed rapid viral spread before verifying content.
Regulatory Environment
Lack of Global Labeling Mandates
Inconsistent international laws fail to enforce AI content labeling.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
