Yeah, all the safety concerns are very valid. Easiest immediate fixes that we're building into our system are 1) preventing people from scaling before doing KYC, 2) creating a content moderation filter on the AI's prompts

Will do more as we grow.