AI-Generated Video Circulates False Claims of Violence in Guadalajara Following Cartel Leader’s Death
A viral video circulating on social media platforms, including Facebook, X, YouTube, Instagram, and TikTok, falsely depicts a scene of fear and unrest in the mountains near Guadalajara, Mexico. The video, which claims to show the aftermath of a military operation targeting “El Mencho,” leader of the Jalisco New Generation Cartel (CJNG), has been identified as entirely or significantly altered by artificial intelligence (AI).
The Viral Video and Claims of Violence
The 15-second video features a young man stating, “Send this video to your son, carnal. Neta, don’t freak out, I was here today. Look how I’m doing: There are no cars, there are no girls, pure fear for 3,000 pesos, you better study.” Some online posts falsely claim the man in the video is a deceased hitman. The video surfaced amidst reports of roadblocks, burning vehicles, and clashes with soldiers in several regions of Mexico following the reported death of “El Mencho” in Tapalpa, Jalisco. Guadalajara and Puerto Vallarta, both in Jalisco, experienced increased violence, raising concerns about potential further unrest during a leadership transition within the cartel.
AI Detection Reveals Manipulation
Analysis of the video revealed several irregularities characteristic of AI-generated content. These include a blurred and distorted background landscape, a blurred and momentarily distorted face of the man speaking, and a lack of visible pupils in his eyes despite adequate lighting. The antenna of his portable radio also appears distorted.
A sequence analysis using the Hive Moderation tool indicated an 85.2% probability that the image and audio were created with AI. Further audio analysis with the Hiya tool InVID-WeVerify returned a 99% probability of synthetic audio.
Expert Confirmation of AI Alteration
Marco Benalcázar, director of the “Alan Turing” Intelligence and Artificial Vision Research Laboratory of the National Polytechnic School of Quito, confirmed that the material “shows signs of having been entirely created or at least altered by AI.” He noted that the man’s teeth appear as a single block at times, rather than showing the natural separation, and that the unchanging black spots representing his eyes are indicators of manipulation.
Tracing the Origin and Spread of the Fake Video
Reverse image searches led to an initial Facebook post on February 25, 2026, at 07:11, featuring the same sequence. Further investigation revealed a Facebook account, “Corridos Perrones,” posting similar videos since February 22, 2026, with young people delivering the same message about a lack of opportunities and the importance of education. Analysis of three of these videos with Hive Moderation showed values over 99% probability of AI generation.
Trend of AI-Generated Misinformation
AFP Factual has identified a growing trend of AI-generated content depicting individuals recording themselves in vertical format videos lasting 15 seconds – the maximum length allowed by platforms like Sora, an OpenAI mobile application for creating AI videos from text prompts. AFP Factual has previously verified other instances of AI-generated misinformation.
References:
- Profile by specialist Marco Benalcázar
- Videos posted on Facebook between February 22 and 25, 2026 (1, 2, 3, 4)
- Account from Facebook “Corridos Perrones” archived