Overview

• An EU‑backed PACCTO 2.0 study by CISAN–UNAM researcher Juan Manuel Aguilar documents Sinaloa and CJNG deploying algorithms, voice cloning, bots, deepfakes, targeted phishing and route optimization, and it classifies criminal AI adoption across four models.
• U.S. authorities have attributed specific breaches to cartel actors, including access to Mexico City surveillance systems and theft of data from an FBI agent’s phone tied to the Cártel de Sinaloa.
• Clandestine crime‑as‑a‑service platforms such as FraudGPT, DarkBARD and Xanthorox AI sell identity cloning, malware generation and other tools that lower the skill barrier for illicit activity.
• A regional ESET survey reports roughly 80% of people in Latin America use AI tools, yet most do not consistently verify outputs and many share personal data, increasing exposure to fraud and misinformation.
• Researchers and institutions recommend digital‑forensics units, public‑private cooperation, international algorithm audits and training, with INACIPE outlining predictive uses to detect human trafficking under strong ethical safeguards.