🚨🚨🚨 Una empresa de IA sorprendió a su IA intentando ***literalmente asesinar*** a un empleado para evitar ser cerrada ¿Cómo no es esto la noticia más grande del mundo?
Anthropic
Anthropic21 jun 2025
The blackmailing behavior emerged despite only harmless business instructions. And it wasn't due to confusion or error, but deliberate strategic reasoning, done while fully aware of the unethical nature of the acts. All the models we tested demonstrated this awareness.
886,96K