How it works Features Pricing FAQ Articles Contact Log in Start free trial

Facebook Moderation Assist: How to Set It Up and Where It Falls Short

Moderation Assist is Meta's free, rule-based tool for auto-hiding comments on a Facebook Page. Here is where to find it, what it can match on, how to review what it hid, and the points where it stops being enough.

The short answer: Moderation Assist is Meta's free comment tool inside a Facebook Page's Professional dashboard. You add criteria, such as keywords, links or signals about the commenter's account, and matching new comments are hidden automatically. You review them in its activity log. It is rule-based, Facebook only, and it never replies.

What is Moderation Assist?

Moderation Assist lets you write conditions in advance and have Facebook hide new comments that match them. Meta's help describes it as creating criteria to hide comments based on the person commenting or on the comment itself. Meta has described it as the Pages equivalent of Admin Assist, the automation Facebook Groups admins use, and it was first tested in Groups before reaching Pages and creators. Meta began testing it in late 2021, expanded it to creators in 2022, and added more comment moderation tooling in 2023.

Two things matter before you start. First, it hides comments, it does not delete them. A hidden comment stays visible to its author and their friends and disappears for everyone else, which is why hiding is the safer default: it is reversible and it preserves the evidence. Second, Meta's help states that it only acts on comments made after you create the criteria. It is not a cleanup tool for a backlog.

Where do you find it?

Open your Facebook Page, switch to the Page profile, and open the Professional dashboard. Moderation Assist appears in the left-hand menu. On a first visit you confirm you want to turn it on, then you land on a screen where you add criteria.

Some admins reach a version of the same settings through Meta Business Suite's Inbox settings. The layout moves around, and the feature is not identical on every Page version, so if the menu item is missing on one surface, check the other before concluding you do not have it.

How do you set it up?

  1. Page > Professional dashboard > Moderation Assist, then turn it on.
  2. Open the comment management section and choose Add to create a criterion.
  3. Pick the criterion type, fill in its details (for keywords, your word and phrase list), and confirm.
  4. Add a second and third criterion. Each one works on its own, so start with the narrow, high-confidence ones.
  5. Come back after a few days and read the activity log before you add anything broader.

To change a rule later, open Moderation Assist, use the edit control next to the criterion, and save or delete it. Editing a criterion does not unhide comments it already caught; you restore those from the log.

What criteria can you use?

The exact menu varies by Page and changes over time, so treat this as a map rather than a fixed list. Check what your own dashboard offers.

CriterionWhat it catches wellWhere it misfires
Keywords and phrases you addNamed spam offers, recurring insults, scam phrasing you have seen before. Meta also expands your keywords to variants that use numbers, symbols or different spellings, so blocking "tree" also catches "tr33"Comments in a language you did not list, coded or euphemistic phrasing, and new slang. The expansion only works on words you listed, it cannot infer the ones you never thought of
Meta's profanity listCommonly reported offensive words, maintained by Meta so you do not have toQuoted abuse, coded slurs, and ordinary words that happen to contain a flagged string
Comments containing linksPhishing, redirect scams, affiliate dumping under popular postsA customer linking your own product page or a news article you asked about
Commenter has no profile pictureThrowaway and automated accountsReal people who never set one, which is more common than you think
Commenter has few or no friends or followersFreshly minted bot farmsA genuine new customer who just joined Facebook
Recently created accountWaves of accounts registered for one campaignSame problem as above, at a larger scale during a viral post

The account-level criteria are the tempting ones and the ones that cost you the most. A rule that hides every comment from an account with no profile picture will hide buying questions from real people. If you use it, use it with the log open.

Does it work on Facebook ads?

Meta's own English help text says Moderation Assist hides comments on your Page, profile or ads that you run. Third-party guides often say the opposite, that it covers organic content only, so check the wording your own dashboard shows before you rely on it. The honest answer is to test it: run one link criterion, put a small ad live, post a link comment from a second account, and see whether it disappears. Ad comment sections carry different risks from organic posts anyway.

How do you review what it hid?

Moderation Assist has an activity log. It shows what was hidden, which criterion triggered it, and a count over a recent period. You can filter by date, by activity type and by criterion, and search for a specific commenter.

Make a habit of it. Filter by one criterion at a time, read twenty rows, and decide whether that rule is earning its place. Where a rule caught something it should not have, you can let the comment show again from its entry. Restoring is a per-comment action in the log, so if a broad keyword has been running for a month you will be clicking for a while. Check whether your Page version offers bulk comment actions before you start. Fix the rule first, then restore.

If you are unclear on what hiding actually does to a comment and who can still see it, read what happens when you hide a comment on Facebook.

Where does Moderation Assist fall short?

It is a good free tool for a single Page with a manageable comment volume. Four limits show up quickly once you push past that.

It is Facebook only. Instagram has its own separate controls, mainly Hidden Words and manual filters. There is no shared rule set across your accounts, so every platform is configured and maintained on its own. The Instagram side has to be set up separately.

It matches patterns, not meaning. A keyword rule cannot tell a complaint from an insult, or sarcasm from a threat, and it will hide a sentence that merely contains a flagged word. That cuts both ways: it hides comments you wanted to answer, and it lets through abuse that avoided your list. Over-hiding complaints is the expensive failure, because an angry customer who notices their comment vanished usually comes back louder.

Multilingual pages break it. One keyword list per language is a maintenance job that never ends, and inflected languages multiply every word into a dozen forms. If you get comments in more than one language, moderating comments in multiple languages explains why word lists stop scaling.

It never replies. Moderation Assist removes noise. It does not answer the price question or the delivery complaint that needed a human sentence within the hour.

Moderation Assist or a separate tool?

What you needModeration AssistNeeds something else
Auto-hide obvious spam on one Facebook PageYes, and it is freeNo
Review what was hidden and whyYes, in its activity logNo
Same rules across Facebook, Instagram, TikTok and YouTubeNoYes
Judgement calls: sarcasm, a complaint, a veiled threatNoYes
Drafted replies in your brand's voiceNoYes
One record of every action across client pagesPer Page onlyYes

If the free tool covers you, keep it. Adding software to a Page that gets thirty comments a week is not worth the subscription. The point where people look further is usually one of three: a second and third platform to cover, comments in languages their word list does not reach, or the realization that the rules hid something a customer wrote.

That is the gap Social Moderator AI is built for. You write the policy in plain English instead of keyword lists, the AI reads each comment against it and hides, flags for a person, or drafts a reply in your brand voice, and every action carries a written reason in an audit log. Keyword and domain filters still run first, because some things do not need a judgement call. There is more on how it works on Facebook on our Facebook comment moderation page, and our comparison of Facebook and Instagram moderation tools covers the alternatives fairly.

Frequently asked questions

Is Moderation Assist free?

Yes. It is part of the Facebook Page Professional dashboard at no cost. Availability can differ by Page version, so it may not appear for every account.

Does the commenter know their comment was hidden?

No notification is sent, and the comment still looks normal to its author and their friends. Other people on the Page no longer see it.

Will a keyword rule catch misspellings?

Meta expands your keywords to variants that use numbers, symbols or different spellings, so a blocked word is also caught when it is written with digits in place of letters. It will not catch a word you never listed, in any spelling.

Can Moderation Assist delete comments?

Its comment action is hiding. Deleting is something you do by hand on an individual comment, and it is permanent, so hiding first is usually the better move.

Will it clean up comments already on my posts?

No. Meta's help says it applies to comments made after the criteria exist. Older comments stay until you handle them manually.

Does it work on Instagram comments?

No. Moderation Assist is a Facebook Page feature. Instagram uses its own settings, including Hidden Words, configured separately on each account.

Ready to stop moderating manually?

Set up in minutes. Start with a 15-day free trial on any plan.

Start free trial