Do AI posts get penalised? What Instagram, TikTok, YouTube and LinkedIn say, and what nobody knows yet
Short answer: As of November 2026, the platform pages cited below do not say that AI written or AI made posts are ranked lower because of how they were made. They do say that realistic AI content must be labelled in some cases, and that failing to label can bring penalties. No study cited here measures an AI penalty. So the answer is that labelling failures can be punished, and a reach penalty for AI origin is unproven. Tools such as Blotato, a software subscription that writes and schedules social posts, do not change that.
You are using AI to write or make posts and worry that Instagram, TikTok, YouTube or LinkedIn will quietly show them to fewer people.
I'm Ranjit, and I run Blur Studio: 15+ years building software, 120 apps taken live, and clients who have gone on to raise $10M+. I build AI systems and processes that help businesses do a lot more with a lot less time. I also work differently from most of what else is out there, more on that further down this page. Say hello.
This question gets confident answers from people selling courses and tools, and few of them have evidence. This page sticks to what the platforms themselves publish, says what is not known, and gives you a way to test your own account. I mention Blotato, which writes, schedules and posts social content, only because many readers use a tool like it. Nobody at Blur Studio has run it, and Blur Studio has no test results to report. This is information, not legal advice. Platform rules change, so check the linked pages.
This page may contain affiliate links from which Blur Studio earns commissions.
Three different questions
People mix up three things, and the answers differ:
- Is AI content banned? No platform page cited here bans it. The rules are about honesty and labelling.
- Must I label it? Sometimes, yes. See below and do you have to label AI content.
- Does the algorithm show it to fewer people because it is AI? Nobody outside the platforms can say for certain. This is the open question.
What the policy pages say
| Platform | What its page says | Reach statement? |
|---|---|---|
| Meta (Instagram, Facebook) | Content may be labelled automatically when industry standard AI signals are detected. People may, and in some cases must, label. Meta may require a label for photorealistic video or realistic sounding audio that was digitally created or altered. The requirement does not apply to images. Penalties are possible for unlabelled content when a label is required. | No ranking penalty for AI origin stated on the page |
| YouTube | Disclose realistic content that is AI generated or meaningfully altered. No disclosure for clearly unreal content or minor edits such as beauty filters, colour, captions or upscaling. Repeated non disclosure can bring penalties including removal or suspension from the Partner Programme. YouTube may add labels itself. | States that disclosure alone does not limit audience reach or monetisation |
| TikTok | TikTok's support page says creators must label all AI generated content that contains realistic images, audio and video, and encourages labelling of content completely generated or significantly edited by AI. TikTok may add an "AI-generated" label itself, for example for content made with its AI effects or carrying C2PA Content Credentials (a standard that attaches origin data to files). Its API has an AI generated content field that adds a label. | States that turning on the AI-generated content setting will not affect distribution of a video that does not break its Community Guidelines |
| No creator AI label rule is cited here. Check LinkedIn's help centre. | None cited |
Read the table closely. Every penalty it lists is for not labelling when required, or for misleading content. None is for using AI as such. Meta and TikTok both tie labels to realism. A cartoon style landscape video is a different case from a realistic video of people in a market.
What studies show
This page cites no study. Vendor blog posts that compare posts by AI and by humans tend to use small samples and are written by companies that sell the tools. Treat any claim of "AI posts get X percent less reach" as unproven unless you can open the method, the sample size and who paid. Likewise treat "AI posts get more reach" the same way. What exists is anecdote, and anecdote mixes up AI origin with quality, timing, topic and account size.
Quality over origin
What platforms say they reward is whether people watch, read, save, share and reply. A flat, generic post gets ignored whether a person or a machine wrote it. A sharp one with a real example, a real photo and a clear point does not. The same is true of repeated, near identical posts across accounts, which look like spam. Blotato's own advice is that bans and rejections come from bot like behaviour, such as posting without ever logging in or engaging, and not from the API route. That is the vendor's view, not a platform statement.
My advice: write the post for a person, add a fact only you know, label where the rule says, and measure your own results.
Risks that are real
- Unlabelled realistic AI video or audio on Meta, or realistic altered content on YouTube.
- Misleading content. A fake customer quote or an invented statistic is a risk under advertising and consumer rules, not just platform rules. See UK rules for social media marketing.
- Tool gaps. Blotato says it does not yet pass Instagram's AI disclosure flag, so a label you need on Instagram must be added in the app after posting. It lists AI fields for TikTok and YouTube.
- Spam like patterns. Same text posted many times, mass tagging, rapid retries.
A test plan you can run
You can find out for your own account. This is a blank plan, because Blur Studio has no results to show. Use posts you would publish anyway, and do not add extra posts or retries to run it.
- Pick one platform and one topic where you post often.
- Make two groups of 10 posts, alternating days: group A written mainly by you, group B drafted by AI and edited by you. Same topic mix, same time slots, same format, same label rules.
- Do not hide anything. Label per the platform's rule in both groups.
- Record reach, views, saves, shares, comments and follows 48 hours after each post.
- Compare the middle value of each group, not the best post.
- Pass: group B is within about a quarter of group A on reach and engagement. Fail: group B is clearly lower. Then find which part: the topic, the hook, the visual, or the labels.
- Repeat after four weeks, since 10 posts is a small sample. If you can only run it once, say so in your notes.
| Date | Group (A or B) | Format | Label used | Reach at 48h | Saves, shares | Notes |
|---|---|---|---|---|---|---|
You can run this with a 7 day trial of a scheduler, or by hand in the platform's app. Its schedule saves effort but changes nothing in the test.
How to read your results
Do not look at one post. Look at the middle value of each group and at the range. If group B is lower on reach but higher on saves, the AI drafts may be clearer but less eye catching, and the fix is the hook, not the tool. If group B is lower on everything, check whether the topics were really equal. If both groups swing widely, the test is too small to say anything, so run another 10 each.
Two traps to avoid. First, do not let the AI group get the better slots. Second, do not label one group differently, because the label itself could change how people react. Keep the variables to one: who wrote the first draft. Write down your own cut off before you start, so you do not pick the answer you hoped for afterwards.
What to do in practice while the question is open
Treat AI as a drafting aid, not an author. Start from your own facts, examples and photos, let the AI structure and tighten, then edit until it sounds like you. Keep a note of which posts used realistic AI video, audio or people, so you can label them quickly if a platform asks. Do not post the same AI text to many accounts. Vary the hook, the length and the format.
If a post suddenly gets far less reach than usual, do not jump to blame AI. Check the timing, the topic, a recent format change, whether a link is in the text, and whether the account got any notice in its settings. Most dips have an ordinary cause. Keep the log from the test plan going, so you have your own numbers to compare when it happens.
What to check each quarter
- Re-read Meta's, YouTube's, TikTok's and LinkedIn's AI content pages and note any changed date.
- Check your tool's AI label fields still work as described.
- Review your own test log.
Who should skip AI drafting, and when a tool is not the answer
If your brand voice is the product, such as a personal expert or creative brand, I would write by hand and use AI only for outlines. If you cannot spend time editing, AI drafts will read generic. If you need an approval trail for client work, Blotato has no approval feature, so pair it with a review step. See a 10 point check before publishing.
How this page was checked
The facts come from Meta's AI content label page, YouTube's help page on altered or synthetic content, TikTok's support and developer pages, and Blotato's help centre, read on 21 November 2026 and refreshed every 90 days. Blur Studio has not run Blotato itself, so test with the 7 day trial plan or an equivalent. Blotato's pricing page is in the sources for plan terms.
How I'd approach this
I would not worry about an AI penalty and would worry about labels and quality. Week one: write down which of your posts contain realistic AI video, audio or people and add the platform label where it applies. Week two: start the 10 and 10 test on your main platform. Week four: read your log and decide how much AI drafting to keep. If you want a second view on your workflow, use the free AI tools or a free AI brainstorm call. If you need a scheduler, Blotato is one option to trial.
I'm not a lawyer, accountant or other licensed adviser, and this page is general information, not advice. Rules and prices differ by country and change often, so treat it as a direction to think along, and please check with a lawyer or the relevant regulator before you act on it.
Questions people ask
Does Instagram show AI posts to fewer people?
Meta's AI label page, as cited here, states labelling rules and possible penalties for unlabelled content when a label is required. It does not state a reach penalty for AI origin. Whether one exists in practice is unknown, so test your own account.
Does YouTube demonetise AI content?
YouTube says disclosure alone does not limit reach or monetisation, and that repeated failure to disclose when required can bring penalties including removal or suspension from the Partner Programme. Read its help page for the current rules.
Does TikTok require an AI label?
TikTok's API has an AI generated field that adds a label, and its support page says creators must label realistic AI generated content. It also says the AI-generated content setting does not affect distribution of a video that follows its Community Guidelines.
Does LinkedIn penalise AI written posts?
No LinkedIn rule on this is cited here. Check LinkedIn's help centre and test your own posts with the plan above.
Are there studies on AI post reach?
None are cited here. Vendor studies often use small samples and are written by tool sellers, so treat them as unproven.
Do I have to label AI images?
Meta says its AI label requirement does not apply to images, while YouTube asks for disclosure of realistic altered content. Rules differ, so see the labelling guide.
Does Blotato add the AI label for me?
It lists AI fields for TikTok and YouTube. For Instagram it says it does not yet pass the AI flag, so label in the app. Check its current help pages.
More in this topic: Social media: rules, safety and compliance (13 guides) →