August 2025 round: 45,000+ prompts. Character.AI called it ToS violation. OpenAI investigating. Google could not confirm ToS breach from available info.