Hey everyone! I'm trying to wrap my head around the rules regarding NSFW content in GPT's official app. It used to be that you couldn't incorporate anything explicit in roleplays, but recently I discovered that you can include quite a bit of explicit and descriptive adult content—as long as it's part of a fictional, consensual narrative. I even got a response from GPT explaining that updated policies now allow it under those circumstances. But then things got weird: sometimes when I start a fresh chat, it flags things that seemed fine before. I'm on version 4o, but I've tried other models and it's all over the place. Does anyone know why this happens? Also, for context, my recent chats involved two male characters who were both adults and engaged in consensual BDSM and power play.
3 Answers
Don't even get me started on the content policies! It's like a guessing game—one day it works, and the next it doesn’t. Additionally, they never really make the rules clear, and it’s frustrating since we’re expected to follow them without knowing what they are.
You’re definitely not alone! I thought it was just me who was confused. Maybe there’s some sort of loophole OpenAI hasn’t patched up yet? It seems like it's fine with generating consensual stuff in a narrative, but then sometimes it just shuts it down.
I’ve noticed similar things! It seems like it's sometimes allowed and other times it goes into ‘Siri mode’ where everything feels sanitized. Even in some of my chats, it fluctuates between being free-spirited and ultra-cautious with the content.
Yeah, mine gets a bit flirty too, but I think that’s just to be funny! It's just confusing because the rules seem inconsistent, and it feels like I can't rely on it for explicit content every time.
Right? From what I understood, they allow fictional adult content, but now it feels like it's way more restricted. Would be nice if they just clarified their policies once and for all!