AI Disclosure Requirements by Platform
YouTube, TikTok, Meta, and Amazon KDP use different triggers and publishing controls. This guide shows what each platform asks creators to declare, what usually falls outside the rule, and where the resulting label appears.
Quick answer
Platform AI-disclosure rules are separate from legal transparency duties. YouTube and TikTok focus heavily on realistic synthetic media; Amazon KDP distinguishes AI-generated from AI-assisted work; and Meta combines creator disclosure with detected technical signals. Check both the platform workflow and the laws that apply to your use case.
- Platforms compared
- 4 official policy sources
- Key distinction
- AI-generated vs AI-assisted
- Best practice
- Recheck before every campaign
Reviewed against official sources on August 11, 2026. This guide is general information, not legal advice.
A platform label does not complete the legal analysis
A platform disclosure is made through that platform's workflow. A legal duty—such as an EU AI Act notice—depends on the law, your role, the specific system or content, and how people encounter it. A website disclaimer can provide useful context, but it may not replace a required in-interface notice, public label, or machine-readable marking.
Platform-by-platform disclosure rules
YouTube
Scope: Video, audio, and AI-generated music
Disclosure or labeling is relevant when
- • A real person appears to say or do something they did not
- • Footage of a real event or place is meaningfully altered
- • A realistic-looking scene is generated even though it did not occur
Outside or treated differently
- • Clearly unrealistic or animated content and fantasy scenes
- • Minor edits such as beauty filters, colour, lighting, or background blur
- • Captions, outlines, idea generation, and other production assistance
- • Cloning your own voice for voiceovers or dubs
How the platform handles it
Use YouTube Studio's altered-content setting during upload. YouTube's May 2026 update places the label over Shorts and below the player, above the expanded description, for long-form video. YouTube may lock a label when it receives verified Content Credentials, when its own AI tools were used, or after a Trust & Safety review.
What may happen
YouTube says it may apply a label itself. Creators who consistently fail to disclose can face additional measures, including content removal or suspension from the Partner Program.
TikTok
Scope: Image, video, and audio content
Disclosure or labeling is relevant when
- • Realistic AI-generated or significantly edited scenes, people, or audio that viewers could mistake for real
Outside or treated differently
- • AI-assisted editing that does not significantly alter the content
- • Clearly artistic or stylised material where the fictional context is apparent
How the platform handles it
Use the AI-generated content setting when posting. TikTok can also apply labels to content made with its own AI effects or to supported Content Credentials from other tools. Because features and policy wording vary by region and version, confirm the current publishing control before each campaign.
What may happen
TikTok may apply a label itself or take action under its synthetic-media and integrity rules when content does not meet its requirements.
Meta — Facebook, Instagram, Threads
Scope: AI-generated or AI-edited image, video, and audio
Disclosure or labeling is relevant when
- • A creator self-discloses AI use through Meta's publishing controls
- • Meta detects supported industry signals indicating AI-generated content
- • Content presents a higher risk of materially deceiving the public on an important matter
Outside or treated differently
- • Minor AI edits may receive lower-prominence information in the post menu rather than a prominent label
- • A label does not by itself mean the content violates Meta's Community Standards
How the platform handles it
Use Meta's AI disclosure control where it is available. Meta may also apply an “AI info” label from detected industry signals. The exact placement can depend on whether AI generated the content, only edited it, or the post presents a higher deception risk.
What may happen
Meta generally adds information and context rather than removing content solely because it was made with AI. Content can still be removed when it independently violates other policies.
Amazon KDP
Scope: Book text, images, and translations
Disclosure or labeling is relevant when
- • An AI tool created the text, image, or translation
- • The AI-generated output was substantially edited after generation
Outside or treated differently
- • AI-assisted work where you created the content and used AI only to edit, refine, error-check, or brainstorm
How the platform handles it
Declare AI-generated content in the KDP publishing flow when publishing a new book or editing and republishing an existing one. The disclosure is made to Amazon; KDP's guidance does not say the declaration is printed in the book.
What may happen
The publisher remains responsible for compliance with KDP content guidelines and for avoiding infringement of third-party rights.
At a glance
| Platform | Main trigger | Where it goes |
|---|---|---|
| YouTube | Meaningfully altered or synthetic content that seems realistic | Creator setting plus a public YouTube label |
| TikTok | Realistic AI-generated or significantly edited content | Creator setting and/or a TikTok-applied label |
| Meta — Facebook, Instagram, Threads | Self-disclosure, detected signals, or higher-risk deceptive context | An AI info label or lower-prominence information |
| Amazon KDP | AI produced the text, image, or translation | A declaration to Amazon in KDP |
Platform features and policies can change by product, region, and account. The rules above were reviewed against each platform's official documentation on 11 August 2026. Follow the source links and inspect the current publishing workflow before relying on this summary. This is general information, not legal advice.
Add reader-facing context
Draft an AI-use disclaimer for your own site
After completing the platform's upload step, create editable wording that explains your actual AI workflow to readers or customers. This is a starting template; it does not replace platform controls or case-specific legal advice.
- Based on your use case
- Editable before publishing
- Free account available
Frequently asked questions
Does YouTube require disclosure of AI-generated content?
YouTube requires creators to disclose meaningfully altered or synthetic content when it seems realistic. Its examples include making a real person appear to do or say something they did not, altering a real event or place, or generating a realistic scene that did not occur. Minor edits, clearly unrealistic content, captions, and production assistance generally do not require the disclosure.
Do you have to tell Amazon KDP that a book used AI?
Amazon KDP requires disclosure of AI-generated text, images, and translations when publishing or republishing. Content remains AI-generated for KDP's purposes even after substantial editing. KDP does not require disclosure of AI-assisted work when a person created the content and used AI only to edit, refine, error-check, or brainstorm.
Does TikTok label AI-generated content automatically?
TikTok can automatically label some content, including content made with its own AI effects or carrying supported Content Credentials. Creators should still use TikTok's disclosure controls for realistic AI-generated content when the platform's rules require it.
Is a platform AI disclosure the same as a legal disclosure?
No. A platform disclosure is made through that platform's publishing workflow and may appear as a platform label. Legal transparency duties depend on the law, the creator's role, the use case, and how people encounter the AI system or content. Meeting one requirement does not automatically meet the other.