How Instagram's Unoriginal Content Policy Works, and What Counts as Original

Instagram's unoriginal content policy is the part of Meta's Original Content Guidelines that removes an account from recommendations when it mostly reposts work it did not make. The current version took effect on 30 April 2026, when Instagram extended rules it had applied to Reels since 2024 to photos and carousels (Instagram, 30 April 2026). Accounts that primarily re-upload other people's work are no longer eligible for recommendations anywhere on Instagram.

Three details get lost in most of the coverage, and all three matter more than the headline.

It judges the account, not the post. Meta's wording is that if your account "primarily posts unoriginal reels, photos, or carousels you didn't create or edit in a material way," the account may stop appearing in recommendations. One weak clip will not sink you. A pattern will.

It runs on a rolling 30-day window, with no published number. The 2024 version counted: accounts that posted other users' content "10 or more times in the last 30 days" lost recommendations (Instagram, 30 April 2024). The current guidelines drop the number and say "primarily", so nobody outside Meta can tell you the exact line.

Your existing followers are not affected. Meta is explicit that the guidelines "do not impact how we show people content from accounts they already follow." What gets cut is discovery, which is the part a brand is usually paying for.

And clips are not banned.

The rule is not about clips. It is about what you did to them.

A material edit adds something the original did not have: new information, context or a point of view. Meta publishes both sides of the line in its Original Content Guidelines, so there is no need to guess. Meta quotes on this page come from that document and from Instagram's announcements of 30 April 2024 and 30 April 2026.

Meta also gives a test, and it is the clearest line in the entire policy.

Read that as a brief, not as a policy. If a clipper's only contribution is a caption and a crop, removing it changes nothing. That clip is a repost with extra steps.

Instagram counts a post as original when you made it, made it with an approved partner, or added something the source did not have; the account is then judged on its last 30 days. Answer five questions about a post you are about to publish. Each flag quotes the rule behind it, and the checker never guesses a threshold Meta has not published.

In short: a straight re-upload, or one with only a caption, subtitles, border, speed change or credit added, does not count as original; narration, commentary, graphics or new context can. An account that mostly posts the first kind loses recommendations until most of its last 30 days are original, and a cut identical to one already posted loses out to the original.

There are two mechanisms here, not one, and almost nobody writing about this has separated them.

The originality rule judges an account over 30 days. Duplicate detection judges a post.

Instagram's 2024 announcement set out how the second one works: "when we find two or more identical pieces of content on Instagram, we will only recommend the original one," and the original "will directly replace the reposted content in recommendations." It does this when the content is relatively new and Instagram is confident the original is on Instagram. Where it does not replace a repost, it adds a label linking to the original creator, visible to the reposter's followers (Instagram, 30 April 2024).

Meta does not publish how close two cuts have to be before it treats them as the same. For anyone running volume, that gap is the whole point.

So a campaign now has two separate requirements.

The first requirement is discussed everywhere; the second barely at all, and for anyone running volume across many accounts it is the sharper risk.

It also ends the laziest cross-posting habit. Meta lists watermarks among the low-effort edits that do not make a clip yours, so a TikTok logo left on a Reel marks it as a re-upload, not an edit. Cut it natively instead, and post it when your Reels audience is actually scrolling; the best times to post on Instagram lays out where Sprout Social, Hootsuite and Buffer agree.

The honest version is uncomfortable, so here it is plainly.

A campaign where 50 accounts post the same file with different captions is exactly what both systems were built to catch. It was already a weak strategy. It is now a measurable liability, because the reach you are paying for is the reach being withheld. That posting pattern is also the clearest sign you are buying from a farm rather than an agency, which we break down in clipping farm vs a real clipping agency.

A campaign where 50 creators each build a genuinely different edit is not the target. Different in-points, different framing, different voiceover, different overlay. Same source, different work.

The difference is not who you hired. It is what they did to the clip, and how different each one is from the next.

Yours probably is not, and this is the part brands get wrong in both directions. The policy penalises the account that posts. In a distribution campaign that is the creator's account, not your brand page. Your own profile is only exposed if you are reposting other people's material to it yourself.

That sounds like good news and it is not, quite. The reach you paid for lives on those creator accounts. If they get limited, your campaign gets quieter and your own Account Status will look perfectly healthy the entire time. You will not see the problem where you are looking for it.

The cost side. There is a second-order effect worth budgeting for. The video agency 3rd + Lamar, writing in May 2026, predicted a 15 to 30% rise in clipper costs entering Q4 as creators price in the extra editing work and the lower earning ceiling on Instagram. That is their forecast rather than an observed figure, and we have not seen a shift of that size yet. But the direction is right. Real transformation takes real time, and time gets billed.

It is worth noting who has stepped forward to answer this so far. In July 2026 the loudest response came from GeeLark, a cloud-phone platform selling automation for multi-account operations. Their own line is that agencies "can no longer rely on near-identical workflows across every profile," which is correct. The answer they are selling is better infrastructure for running many accounts. The answer Instagram is asking for is different content on each one.

Meta has published two carve-outs. Almost no coverage mentions either, and one of them is easy to misread badly.

The one you can use: the branded content tool. Meta's own instruction is to share content you produced yourself "or created with approved partners (e.g. content with partners as part of our branded content and partnership ads tool)."

That is a real route, and it is open to any brand. The mechanics are rarely published, so here they are.

Both sides need a professional account, Creator or Business. Instagram expects an authentic, established presence with a reasonable follower base. The creator requests permission to tag you, you approve it, and only then does the "Paid partnership with" label appear on the post. Once branded content is running, partnership ads let you amplify a creator's organic post with paid targeting behind it.

If your agency tells you clips are disclosed, ask which method. The answer should be the tool, not a hashtag.

There is also a floor to get over. Instagram expects a real account with a reasonable following on both sides, so a campaign built on brand-new accounts with a hundred followers each will not clear it. If that is your situation, the tool is not available to you yet and material editing is the only route you have.

What about Collab posts? Meta now names them. Its 30 April 2026 update lists "collabs, remix, or the paid partnership label for branded content" as ways to include third-party content in posts you had a role in creating, while everyone involved gets credit (Instagram, 30 April 2026). Collab posts put two accounts as co-authors on one piece of content, and the post appears on both profiles.

That settles the argument 3rd + Lamar and others made earlier in the year. It does not make Collab posts easy at volume: every one needs the other account to accept an invite, which is manageable across five creators and painful across five hundred.

The alternative Meta actually suggests. If you genuinely want to put someone else's content in front of your audience and you are not going to change it, Meta names two safe options: repost it to your followers using the built-in repost tool, or share it to your stories. Both credit the original creator and neither counts against you. What is not safe is posting it to your grid as though you made it.

The one you cannot rely on: the publisher exemption. Instagram's 2024 announcement said the aggregator rule "will not impact a set of publishers we've identified who have licensing agreements and/or explicit permissions from content creators" (Instagram, 30 April 2024). The current guidelines do not repeat that line.

Read that carefully. It says a set of publishers we've identified. That is a curated list Meta maintains. It is not an automatic exemption for anyone holding signed permissions.

An agency with creator agreements in a drawer is not covered by default, and any vendor who tells you otherwise has misread the sentence. We hold agreements with every creator in our network and we still treat the originality rules as fully applicable to us, because they are.

One last thing worth separating. Meta states that these guidelines are "separate from our intellectual property policies." Originality and copyright are two different systems with two different consequences. If the copyright side is what you are worried about, that is covered in our guide to is clipping legal.

"Shadowban" is not a thing Instagram documents. "Limits to your reach" is, and you can check it.

Account Status, in your settings, shows whether your account has any limits to its reach. A reach limit from unoriginal content means the account is not eligible to be recommended to people who do not follow it.

Recovery works the same way the penalty does. Meta says eligibility returns "when most of your recently posted reels, photos, and carousels are considered original in a 30-day period," calculated on a rolling basis. You have two options: remove the unoriginal content, or appeal if you think the call was wrong.

What Meta does not give you is how long that takes in practice. We could not stand a figure up from a source worth citing, so this page does not print one. Plan on the rolling window above, not on a number somebody guessed.

If a partner account gets limited mid-campaign, the sequence is straightforward.

Most campaigns fail at the brief, and this policy makes a vague brief expensive. Seven things belong in yours. Regulated advertisers have a second reason to plan their overlays, because gambling clips carry mandatory on-screen elements, see gambling clip disclosure requirements.

The workflow for building clips has not changed, and how to clip on Instagram without losing reach walks the native clip action this policy rewards. What changed is that the compliance layer now sits on top of it. On a dated campaign that layer becomes a milestone: the AI product launch marketing plan checks account eligibility at T-3 weeks, because posting history cannot be produced in the final fortnight.

These questions are specific to originality. For the full buying framework, see how to choose a clipping agency.

What transformation is required before a clip ships, and who checks it?

How do you make sure two accounts do not post a near-identical cut?

What is your rejection rate, and what are the top three reasons clips get rejected?

Do these accounts post anything besides my clips?

Are you using the branded content tool where it applies, or just #ad in the caption?

What happens when an account gets reach-limited in the middle of a campaign?

An agency that cannot answer the second and third questions is not running a review process. It is running a queue.

The rejected 56% is the work. Those are the clips that did not clear the bar:

We run 62,900+ creators and have delivered 18B+ verified views, and the approval gate is the reason both numbers mean anything. A model that approves everything is the exact model this policy was written to stop.

If you want to see how that runs in practice, that is what our clipping campaigns and short-form video distribution work is built around, including on Instagram Reels.

That is a problem if you are buying volume. It is an advantage if you are buying work. Lumina Clippers is a clipping agency that distributes brand videos through 62,900+ clippers.