In my latest op-ed for TIME, I discuss why the OpenAI Hugging Face cyber incident represents a turning point for AI safety.