Academy / AI Settings
AI Guardrails: Hard Limits, Coverage and Escalation
Persona and rating set the tone. Guardrails set the edges: what your AI never does, when it is actually on, and when a conversation goes to a real person.
Six AI guardrails are locked on for every creator and cannot be turned off. You can add up to 20 of your own hard limits on top. Coverage decides your AI's working hours, either 24/7 or a daily window. Two conditions stop your AI on a conversation, and three route it to the CreatorLabs team instead.
All three guardrails live together under AI Config. They are the settings worth getting right before you leave your AI running unattended.
Which hard limits are always on?
Six are on by default, locked, and cannot be turned off. They apply to every creator on CreatorLabs with no setup required.
These stay enforced at every content rating, including X. Changing your preset does not loosen them.
How do you add your own hard limits?
Type a rule, hit add, and it appears as a removable tag. A creator who does not want their AI mentioning other creators adds one line:
You can add up to 20 limits in total. Keep each one specific. A rule like "no off platform contact" is enforceable because it is unambiguous. A vague one like "be careful" is not, because your AI has nothing concrete to check against.
If it is more of a vibe than a rule, it belongs in your Creator Context instead.
Does your rule belong in Hard Limits?
- It can be judged true or false on a single message
- It names a thing, not a feeling
- You would want it enforced even on your best fan
- Breaking it once would be a real problem, not just off brand
When is your AI actually working?
Two ways to set your hours.
24/7 Always Online
Your AI is always available. Maximum fan engagement.
Peak Hours
Active during a daily window you define, for example 8 AM to 2 AM, and off the rest of the time.
Coverage also shapes the timing of automatic posts, which are scheduled inside your active hours.
When does your AI stop handling a conversation?
Two conditions pause your AI on a conversation entirely.
When does a chat go to the CreatorLabs team?
Escalation is different from stopping. These conditions route the conversation to the CreatorLabs internal team, who step in directly.
This is the safety net for anything that should not be left to automation. An upset fan, a verification request, anything that brushes against platform rules.
| Setting | What it controls | Who can change it |
|---|---|---|
| Locked limits | Six rules your AI never breaks | Nobody. Always on |
| Custom limits | Up to 20 of your own rules | You, any time |
| Coverage | The hours your AI is active | You, any time |
| Stop handling | Pauses your AI on one conversation | Automatic |
| Escalation | Sends the chat to the CreatorLabs team | Automatic |
Both cards live under AI Config. Edit, then hit Save. Cancel discards whatever you just changed.
FAQ
How many hard limits can I add?
Up to 20 in total. The six locked limits are included in your view but cannot be removed or edited.
Can I turn off the locked limits?
No. They apply to every creator on CreatorLabs at every content rating and cannot be switched off.
What is the difference between stopping and escalating?
Stopping pauses your AI on that conversation. Escalating hands the conversation to the CreatorLabs team to handle directly.
Does Peak Hours coverage stop automatic posts too?
Posts are scheduled inside your active hours, so a narrower window narrows the times a post can go out.
Why is my custom limit being ignored?
Usually because it is too vague to check against a message. Rewrite it so it names a specific thing your AI must not do.
Next up
Not set up yet?
Connect your page and we walk you through the whole configuration in one sitting.
Start free