What Meta says about original content, and why it affects your reach

Meta publishes, on pages open to anyone, what it considers original content, what it shows less of in the feed and what leads an account to be dropped from recommendations. If you approve brand videos, you have a lot to gain from reading these texts carefully. They contain expressions that should be pinned to the wall of every communications team: "authentic stories," "unique perspective," "recycled content."

In this second article of the series, we go through these official pages from Facebook, Instagram and the Meta Transparency Center with you. We want to know exactly what the company says, so that the conversation about generic AI scripts on social media starts from verifiable facts.

What Meta calls unoriginal content on Facebook

In July 2025, the Facebook for Creators blog published the post "Combating Unoriginal Content to Protect and Elevate Creators." The definition is direct: unoriginal content is content that repeatedly reuses another creator's material without giving credit. Accounts that persist in this practice temporarily lose access to monetization programs and, according to the page, "will also have reduced distribution on everything they share." When the systems detect duplicate videos, distribution of the copies is reduced.

The same page includes figures for the first half of 2025. Meta took action against about 500,000 accounts involved in spam or fake engagement and removed about 10 million profiles impersonating major content creators.

In March 2026, according to Business Standard, Meta revised the Facebook rules and detailed what counts as original: videos filmed or produced by the owner of the page or profile. Simple reactions, "watch along" videos, compilations of clips and voiceovers on other people's videos without added value may be treated as unoriginal and shown less. Repeat offenders may lose eligibility for recommendations and monetization, with the right to appeal.

Instagram: originality as a recommendation criterion

On Instagram, the discussion reached recommendations, the part of the app that takes a post to people who do not yet follow the account. In April 2024, the official page "Helping Creators Find New Audiences" announced that original content would replace the copied version in recommendations and that a new ranking signal would give more distribution to smaller creators. Among the eligibility criteria is content that is "original, does not violate the community guidelines and has no visible watermarks."

In April 2026, Instagram extended the rule to photos and carousels. The new page defines original content as work "that you created entirely or that reflects your unique perspective." A border, watermark, caption or credit is not enough to turn someone else's post into original content. Accounts that mainly publish unoriginal content stop being recommended, and people who already follow them keep seeing their posts. Eligibility returns when most of what was published in the last 30 days is original. Meta reported that 75% of recommendations in the United States already come from original posts.

A 2023 tips page sums up the guidance in a few words: "post original content, avoid reposted content."

Distribution guidelines: what Meta demotes in the feed

Since September 2021, Meta has maintained its Content Distribution Guidelines, which cover, in the words of the official announcement, "problematic or low-quality content." The stated goal is to encourage creators to invest in high-quality, accurate content. In the Transparency Center, the page on domains with limited original content includes a simple sentence: "people have told us they want original content."

One of the demoted categories matters to us, who write scripts, and to you, who approve them: engagement bait. According to Social Media Today, the guideline describes posts that explicitly ask for votes, shares, comments, tags or likes. The same list includes low-quality videos and clickbait links. Picture a Reels script that ends with "comment YES if you agree" or "tag a friend who needs to hear this." This kind of ending, common in formula-driven texts, comes close to what Meta describes as bait.

The Transparency Center also has a page in Portuguese, "Tipos de conteúdo que rebaixamos" (Types of content we demote). Its body did not load during the research for this series, so we do not cite the list here.

And what about artificial intelligence in Meta's rules?

The answer calls for care. The Facebook and Instagram pages on original content cited above do not mention AI. No official rule we checked states that a text written with AI has its reach reduced for that reason.

Meta deals with AI on another front, the "AI info" label, which appears in Portuguese as "Informações de IA." It was announced in April 2024 by Monika Bickert, vice president of content policy, and appears on images, videos and audio, based on industry signals or self-disclosure by the person posting. In September 2024, the company moved the label to the post menu when AI only edited the content. Text does not receive this label, and no source we consulted says it reduces distribution.

If you record your own video with a script written by AI, it fits the definition of original described by Business Standard in March 2026. The risk lies in what generic text tends to produce: the same structure from one post to the next, catchphrases that could appear on any profile, calls that ask for engagement and no story that reflects your perspective.

How to read Meta's guidelines with a script in hand

The official pages point in the same direction. Meta wants to show more people content that comes from the person posting, that adds something to what already exists and that tells a story authentically. The words it chose, "authentic stories" and "unique perspective," describe exactly what a script built from a generic prompt lacks.

If you run a department or a brand, our practical reading is this: the rule probably does not punish a video for using AI, and a video that looks like thousands of others gives the recommendation system little to highlight. People who follow your account also find little reason to stop, comment or share. The third article in the series explains why platforms deliver less of what is generic and repetitive, and the seventh follows what the head of Instagram has been saying about authenticity.