From Deepfakes to Dark LLMs: 5 use-cases of how AI is Powering Cybercrime

This Group-IB report analyzes the growing adoption of generative AI in cybercrime, detailing five primary use-cases—impersonation and live deepfakes, AI-assisted scam call centers, self‑hosted Dark LLMs, AI-powered spam/mailers, and AI-integrated criminal tooling—highlighting multimillion-dollar frauds, underground markets, and operational trends; it concludes with defensive recommendations such as layered detection, underground monitoring, AI-enabled defenses, and enhanced training.