According to Jin10, OpenAI is working to introduce Safety IDs that users can provide during investigations to help distinguish between different end users.