An AI model named MechaEpstein-8000, fine-tuned on the publicly released emails of Jeffrey Epstein, has rapidly gained traction on Hugging Face, exceeding 33,000 downloads. Developed by AIfredo Ortega, the model uses Qwen3-8B to simulate Epstein's communication style, including his dismissive tone and characteristic phrases.
Ortega states the bot is for entertainment and persona mimicry, not research. It runs locally and replicates Epstein's focus on power, connections, and evasiveness regarding wrongdoing. The model's design, including its use of terms like "goy" and its simulated recruitment tactics, reflects patterns found in Epstein's correspondence.
The model's popularity, far exceeding other fine-tuned AIs by the same developer, highlights a public fascination with controversial figures and their digital footprints. This raises significant ethical questions about AI development, dataset usage, and the potential for simulating and disseminating harmful narratives.
