Platform rules on AI-made content and labels
You've used AI for a post and now you wonder whether it needs a label. The pages below talk mostly about images, video and audio, and several single out content that looks real. They don't set out a label rule for AI-drafted text. Each platform words it differently, and the pages change. All were read on 8 November 2026.1
These are platform policies, not laws. Laws on AI disclosure differ by country. This page doesn't cover them.
Facebook, Instagram and Threads (Meta)
Meta announced in April 2024 that "AI info" labels would reach a wider range of video, audio and image content.1 A label can be triggered when Meta detects industry standard AI image indicators. It can also follow when a person discloses that they're uploading AI-generated content.1
Meta describes labelling with extra context as the better route for this content.1 The post was first published in April 2024 and updated afterwards, so Meta's Help Centre has the current steps.
YouTube
YouTube asks creators to disclose AI-generated or altered content that seems realistic.2 It adds that realistic AI content and meaningful changes call for disclosure.2
It also lists things creators needn't disclose. Examples include using generative AI for an outline, script, thumbnail, title or infographic, and caption creation.2
TikTok
TikTok's Community Guidelines require creators to label AI-generated or heavily edited content showing realistic-looking scenes or people.3 The version read here took effect on 24 September 2026.3
Unlabelled content may be removed, restricted or labelled by TikTok's own team.3
LinkedIn said on 15 May 2024 that members would see a new icon for tracing where AI-created media came from.4 The icon shows whether content was created or edited by AI.4 That page sets no rule for businesses writing text posts. Check LinkedIn's own help pages for current terms.
A quick view
| Platform | What the saved page says | Page date |
|---|---|---|
| Meta (Facebook, Instagram, Threads) | "AI info" labels on video, audio and image content | April 2024, updated later1 |
| YouTube | Disclose realistic AI-generated or altered content | Read 8 November 20262 |
| TikTok | Label realistic AI-generated or significantly edited content | Effective 24 September 20263 |
| Icon showing the origin of AI-created media | 15 May 20244 |
What to do in practice
Keep a note of when each post used AI for a picture, video or sound. Suppose you add an AI picture of a scene that looks real, like a product shot that never existed. Use the platform's disclosure setting if there is one.
Don't use AI to show something that didn't happen as if it did, such as a made-up customer or a fake before-and-after. That can mislead customers whatever a label says. Consumer rules worth reading are named in the post checklist. Rules change. Check twice.
X, Pinterest, Google Business Profile and other places aren't covered here. Read their current help pages before relying on any of this. For the wider picture, see what the tools can do.
Where the facts come from
- Our Approach to Labeling AI-Generated Content and Manipulated Media, Meta
- Disclosing use of GenAI content, YouTube Help (Google)
- Integrity and Authenticity, Community Guidelines, TikTok
- LinkedIn adopts C2PA standard, LinkedIn
Also worth a look
What AI can and can't do for small business social postsWhere an AI writing tool helps with social posts, where it needs a person, and who stays responsible for what gets posted.
Giving an AI tool your business's voiceWhy AI posts come out generic and the fixes to try in order, from clearer instructions to real examples of your own writing.
A simple weekly workflow for drafting posts with AISteps from a rough idea to a checked post ready to schedule, with a person approving each one before it goes out.