A rule tells AI both what to do and when to stop
The configuration turns brand strategy into specific rules that can be checked before content is generated or published.
A channel focused on inspiration
Instagram should showcase the stay experience three times a week, without publishing unconfirmed prices or automatically approving new campaigns.
Configuration ready for testing
AI creates content samples, flags breaches and shows the team how the rules will work before the version is approved.
- permitted topics and formats
- mandatory elements and prohibitions
- approval and autonomy thresholds
Strategy turned into operating rules
The module stores versioned rules for channels and content types. It covers objectives, language, frequency, CTAs, restrictions and decisions requiring human involvement.
Organises the rules
Brings scattered decisions about channels, content and approval into one configuration.
Sets boundaries
Defines prohibited topics, unconfirmed promises and cases requiring the process to stop.
Tests the configuration
Shows samples and breaches using historical or example posts before a new version of the rules is published.
How are publishing rules created?
Define each channel’s role
The team defines the audience, objective and desired action separately for each profile.
Set the rules
It defines frequency, languages, formats, mandatory elements, prohibitions and exceptions.
Test with content
AI assesses example posts and shows how the configured rules affect the result.
Publish the version
An authorised employee approves the version, which then applies to generators and automations.
Every rule has an owner and a version
Changing the objective, frequency or level of autonomy can affect multiple modules. The system retains the author, date, rationale and previous configuration version.
What does AI use?
- objectives, audiences and the role of each channel
- frequency, languages, CTAs and formats
- excluded topics and essential sources
- approval levels, limits and exceptions
What does the team receive?
- approved configuration for generators
- content samples and breach list
- versions with change history
Only people can expand autonomy
AI can help describe or test a rule, but cannot independently relax a restriction or increase its own permission level.
An authorised employee approves every new version, sets exceptions and decides which types of content require approval.
The module does not create calendars, generate final assets, approve its own rules or publish content.
Consistent rules across the entire process
Generators and automations use the same approved guidelines.
The team knows what AI cannot do and which cases always require a decision.
The level of autonomy can be expanded in stages, based on testing and version history.
Rules govern the entire publishing cycle
The Central Publishing Panel stores the configuration and its versions. A context interview and additional sources can help identify existing brand and team standards.
Related process elements
Frequently asked questions
Can rules vary between channels?
Yes. Each channel can have its own objective, language, formats, frequency and required approval level.
Can specific topics be blocked?
Yes. The configuration can include topics, claims, wording and types of data that AI cannot use.
Who can change the level of autonomy?
Only an authorised team member. AI cannot expand its own permissions.
Can new rules be tested?
Yes. They can be checked against samples and historical publications before approval.