Short answer: no platform we checked says it ranks a caption or reply lower because AI helped write it. Every one of them punishes the same few things: copied content, spam, unlabelled fake-looking media and bots nobody asked for. Use AI to say something real, quickly, and you're on the right side of every rule.
What we checked
We read the official ranking explainers, community rules and developer policies of seven networks in October 2026. No blog rumours, no "an insider told me". Every rule below links to the platform's own page.
Platform by platform
- Instagram: ranking looks at how many people like a post and how quickly people like, comment, share and save it (Instagram). What it tells creators to avoid is reposted content and watermarked videos from other apps (Instagram). AI isn't mentioned as a ranking factor.
- Facebook: Facebook demotes duplicate and lightly edited copies of other people's posts, and can stop recommending accounts that keep doing it (Meta, 2026). Comment length and relevance are ranking signals (Meta).
- Threads: one of the things Threads predicts is how likely you are to reply to a post (Meta).
- TikTok: realistic AI scenes or people need a label, and reused content with nothing new can be kept out of the For You feed (TikTok). Text isn't covered by the label rule.
- YouTube: "Disclosing AI content alone won't limit a video's audience." Scripts, titles and captions written with AI need no disclosure (YouTube). Mass-produced, template-made AI videos are what lose monetisation (YouTube).
- LinkedIn: bots that comment, like or share are banned, and repetitive comments count as spam (LinkedIn, LinkedIn Help).
- X: automated replies are only allowed to people who asked to hear from you. Replying automatically to keyword searches is not (X).
The pattern
Read side by side, the rules say the same thing: platforms reward content and conversations people want, and punish shortcuts that fake them. How a sentence was drafted isn't on the list. What you do with it is.
- Safe: AI drafts a caption, you add a real detail and post it.
- Safe: AI drafts a reply to a customer's question, you approve it and it answers them.
- Risky: posting the same AI caption on 20 accounts, or reposting other people's videos with an AI voiceover.
- Against the rules: bots that comment on strangers' posts or DM people who never contacted you.
Does AI help reach, then?
Not by itself. In a Buffer study of 1.2 million posts, AI-assisted posts had higher median engagement than the same accounts' other posts. Buffer's own explanation is that people using AI post more consistently, not that AI text is favoured (Buffer, 2024, vendor data). That matches the rules: the lift comes from showing up more and replying faster.
How to use AI the way the rules reward
- Give it your facts, so replies are specific and correct. Accurate, useful answers are what earn more comments.
- Reply to real people who wrote to you. Never use AI to start conversations with strangers.
- Keep posts original: your photos, your news, your offers.
- Label realistic AI images or video where the platform asks you to.
- Approve replies until you trust them.
Read next: Do replies boost reach? and Is automation allowed?
Frequently asked questions
Does Instagram penalise AI-written captions?
Instagram's ranking explainer doesn't list AI as a factor. It warns against reposted and watermarked content, and ranks on likes, comments, shares and saves.
Do I lose reach if I label AI content?
YouTube says disclosure alone won't limit a video's audience. Meta and TikTok labels cover realistic images, video and audio, not captions.
What kind of AI use does get penalised?
Mass-produced or copied content, spam comments, unlabelled realistic fakes and bots that contact people who never asked.
AI content, scheduling, engagement and a full marketing suite — free to start.
Start free
