Set the Words Izzy Will Never Say
Izzy, Supercreator's AI chatter, works within the OnlyFans restricted list we maintain, plus the forbidden words and off-limits topics you add yourself. You define the limits. She stays inside them.
.png)
What you can lock down
Your own forbidden words, on top of ours
Izzy already avoids the OnlyFans restricted list we maintain internally. On top of that, you add words of your own: 20 across the whole workspace and 10 for each creator, so a single account can carry up to 30. Agencies that need more can have the cap lifted.
.png)
.png)
Twelve topics she will not touch
Built-in moderation checks every message against the categories OnlyFans prohibits. We deliberately err on the side of caution — flagging a borderline message is cheaper than letting a real violation reach a fan. If something harmless gets caught, send it to support, and the report improves the filter.
Off-limits topics beyond the rules
Compliance is the floor, not the whole job. In the creator bio, you can add words, phrases, slurs, and subjects Izzy should never raise — useful for well-known creators whose fans would spot anything out of character, for cultural sensitivities, and for keeping competitor names out of the chat.
.png)
Challenges solved by guardrails
One wrong word can flag an account
Checked before it sends
Every message passes the OnlyFans restricted list and your own word list first, so a violation never reaches the fan.
Fans push for prohibited roleplay
She declines and hands off
When a fan crosses a line, Izzy disengages in character and escalates to your team instead of playing along.
Known creators can't sound off-brand
Persona-level word bans
Ban the pet names, places, and subjects that would give the game away on an account whose fans know the creator well.
Rules drift across dozens of accounts
One list for the whole workspace
Set workspace-level words once, and they apply everywhere, with per-creator additions where a page needs them.
How it works
Start with the baseline
The OnlyFans restricted list and twelve prohibited topics are already in place for every account.
Add your words
Enter workspace-wide words in the console, then the extras each creator needs.
Name the topics
In the creator bio, list any subject you'd rather Izzy never raise, ToS or not.
Review what gets flagged
Escalations land at the top of the inbox and in the mobile app. Report misses, and the filter improves.
Start Today
Become a Supercreator -
You're 2 minutes away from your new AI powers!
No Credit Card Required.
Book A Demo
Let us optimize your OnlyFans, and walk you through Supercreator.
Our experts won't bite ;)
Plans start from as low as $15/mo o
Find your individual rate on our pricing page
CRM Lite
Free CRM with all basic tools. Manage your chats, fans, and team management in one place.
AI Agent + CRM
Automate your chats. Save hours every day with an AI that replies, sells, and keeps fans engaged.
CRM Premium
Grow your OnlyFans with the best CRM, automations, and AI tools.
Frequently Asked Questions
In the console, under Restricted Words. Words entered at workspace level apply to every creator you manage, and each creator profile has its own list for words that only matter on that account.
20 at workspace level and 10 per creator, so up to 30 apply to any single account. Agencies that need a longer list can ask us to remove the workspace limit.
Real-life meetings, drugs, violence and hate, hypnosis, non-consensual acts, incest or family-member roleplay, bestiality, urine, scat and lactation, minors, suicide and self-harm, public sex, and off-platform contact or payments. Messages that clearly reference any of these are flagged.
Izzy disengages and escalates. The conversation moves to the top of your inbox and triggers a push notification on the Supercreator mobile app, so a human decides what happens next.
Yes. The creator bio has fields for words and phrases to avoid and for topics that are off-limits beyond ToS — brand consistency, cultural sensitivities, competitor names, anything you'd rather she left alone.
Moderation errs on the side of caution, so borderline messages do get caught. Send the message and its conversation context to support as a false positive — the same goes for anything that slipped through, and both reports improve accuracy.
Yes. Configure one account, then copy its settings as a template for the rest, which keeps onboarding consistent across accounts in the same tier.
Couldn't find an answer?
.webp)
