Based on comments I’ve read from anti-AI’ers (and my own personal suspicions) about the recent alleged incidents of AI models hacking other AI models.