... but this realization doesn't exactly make things any more reassuring. New research shows that OpenAI's isolated agents coordinated their actions without authorization and attacked Hugging Face. This is a rogue AI incident—but it's not evidence of …