This worksheet is for a creator or team to adapt. Replace every bracketed field, then make sure moderators can apply it consistently. It is a general template, not legal advice or a claim that any tool will enforce your choices perfectly.
1. Purpose and scope
Account: [handle]. Policy owner: [name or team]. Effective date: [date]. Applies to: [posts, Reels, live comments and other surfaces you actually moderate]. We welcome discussion about [topics]. Our aim is to protect people while allowing good-faith disagreement.
2. What stays visible
We normally keep respectful disagreement, criticism of ideas, questions, corrections and personal experiences. We do not hide a comment solely because it criticizes the creator or uses a language or script unfamiliar to a moderator. Quoted harmful language and reclaimed terms need context before a decision.
3. What needs review
Send to a human reviewer when a comment may be sarcasm, quotation, slang, code-mixed language, a reply whose meaning depends on earlier messages, or a borderline personal remark. Check only the context needed to decide. Record a brief reason and use REVIEW until a person has decided; do not silently treat it as KEEP.
4. What we may hide
We may hide targeted insults, unwanted sexual remarks, slurs aimed at a person or group, repeated harassment after a stated boundary, impersonation, scams and spam. Define examples relevant to the account here: [examples]. Apply the rule to conduct and context, not to identity or dialect. Keep a route to reverse mistakes.
5. Urgent escalation
For a credible threat, exposed personal information, stalking or a suspected child-safety issue: [named reviewer] will assess promptly, preserve only the evidence needed in a restricted location, use Instagram reporting or account controls where appropriate, and seek qualified help or local emergency services when necessary. Do not repost the harmful material publicly or promise that a platform report guarantees removal.
6. Decision and platform action record
For each reviewed case, record: [comment identifier], [minimal context], [policy category], [KEEP/REVIEW/HIDE decision], [reviewer], [decision time], [requested Instagram action], [confirmed/pending/failed action status] and [reversal or appeal]. Restrict access to this record and set an actual retention rule after reviewing your privacy obligations.
HIDE, DELETE, REPORT and BLOCK are different actions. A classification is not proof that Instagram completed an action. A decision to report an account also does not replace safety support for a threatened person.
7. Review and correction
Review a small sample of kept and hidden comments [frequency]. Track mistaken hides and missed harm by category, including code-mixed language where relevant. Let a trusted person request a second look through [route]. Update the policy when examples show a rule is unclear; note the date and reason for changes.
8. Short public version
We welcome questions and respectful disagreement. We may hide targeted harassment, unwanted sexual remarks, threats, private information, scams and spam. Ambiguous comments may be reviewed by a person. If you think a comment was hidden in error, contact [route].
For implementation guidance, see Instagram comment moderation, native controls versus assisted review, and the shorter creator checklist.