Guide, updated May 16, 2026 · 7 min read

How to disclose AI music on YouTube, TikTok, and Instagram

Each platform has its own AI disclosure rule. The labels live in different places, the policies say slightly different things, and auto-detection is not reliable on music. This guide walks through where to set the flag on YouTube, TikTok, and Instagram for an AI generated song, and what each platform's policy actually requires.

To disclose AI music on YouTube, set the altered or synthetic content flag in the upload form's content details section. On TikTok, toggle the AI generated content label inside the post composer before publishing. On Instagram, enable the AI labeling option in the Reels composer during upload. All three are creator-set, not auto-detected reliably for music, and missing them on a monetized account has produced enforcement actions on real channels. The disclosure does not block monetization, but failing to set it can.

AI music creators publishing across YouTube, TikTok, and Instagram navigate three different disclosure systems with three different labels and three different policy texts. The systems share a goal (label realistic AI generated content for viewers and platform integrity) but they diverge on detail. YouTube uses an altered or synthetic content disclosure in the upload form. TikTok has a manual AI generated content toggle in the post composer. Instagram and Meta use an AI labeling system that depends on the upload path. This guide walks each one, with the actual location of the setting, what the policy text says, and what happens if you skip it.

Before you start

  • A song generated with an AI music tool (Suno, Udio, or any model-driven generator). The disclosure rules apply to AI generated music, regardless of which tool produced it.
  • Accounts on whichever platforms you plan to publish to: YouTube channel via Google account, TikTok account, Instagram account.
  • Awareness that the disclosure flag is your responsibility as the uploader, not the AI tool's. Suno, Udio, and Dayvid do not set the flag for you on the platform side.

AI disclosure rules across YouTube, TikTok, and Instagram

StepPer-platform ruleWhat the creator does
YouTubeAltered or synthetic content disclosure required when content realistically depicts events, places, or people that did not occur, or features realistic AI generated audioToggle the altered content flag in the upload form's content details section
TikTokAI generated content label required for realistic AI generated or significantly edited mediaEnable the AI generated content toggle in the post composer before posting
Instagram and MetaAI labeling required for AI generated or substantially altered media. Some labeling is automatic based on metadata, some is creator-setUse the AI labeling option in the Reels composer when posting AI music videos
Auto-detection on musicLimited and unreliable today. Music is harder to auto-detect than visual deepfakesSet the flag manually on every upload, do not rely on automatic detection
What happens if you skip itStrikes, demonetization, or content removal in extreme cases, depending on platform and severitySet it. There is no upside to skipping
What it does to reachDocumented reach impact is minimal in the typical caseThe label does not visibly tank distribution for music in most reported cases
Where the rule livesEach platform publishes its own AI policy page in its help centerRead the policy page on the platform you are publishing to before publishing

1YouTube: set the altered or synthetic content flag

When you upload a video to YouTube, scroll the upload form to the content details section. There is a question about whether your content contains altered or synthetic content that realistically depicts events, places, or people, or that features realistic AI generated audio. For an AI generated song from Suno or Udio, answer yes. The flag is on the video, not on your channel. Set it on every upload that includes AI music. If you publish through Dayvid, the publish flow sends the video to YouTube Studio as a private draft and you can confirm or edit the metadata, including the AI disclosure, before flipping the visibility to public.

  • YouTube's AI disclosure rule applies to realistic AI generated audio. Most AI music with vocals falls under it. Pure instrumental AI music is a grey area in the current policy text, and the safer default is to disclose anyway.
  • The flag is independent of monetization rules. Setting it does not disqualify your video from YPP. Not setting it on content that needs it can cause issues during review.

2TikTok: enable the AI generated content label in the composer

Open the TikTok post composer with your video selected, scroll the upload options, and find the AI generated content toggle. Enable it for any post that includes AI generated music. The label then appears on the post for viewers. TikTok requires this for realistic AI generated or significantly edited media, and music falls within that scope for the same reason it does on YouTube: it is realistic audio, generated by a model, indistinguishable from human-produced music for most listeners. Setting the label is a per-post action, there is no account-wide default to assume.

  • TikTok's AI detection has improved on visual deepfakes but remains weak on music. Do not assume the platform will auto-label your post; you set it yourself.
  • The label appears as a small indicator below the post. Reports from creators indicate this does not noticeably affect For You distribution for music content.

3Instagram and Meta: use the AI labeling option in Reels composer

When you upload a Reel containing AI generated music, the Instagram composer offers an AI labeling option. Enable it for the upload. Meta's policy covers AI generated and substantially altered media across Facebook, Instagram, and Threads, and the labeling flow lives in the composer per post. Auto-labeling exists when the upload contains C2PA metadata or other provenance signals, but reliability is partial. For a song from Suno or Udio rendered into a Reel through any video tool, the creator-set flag is the safe path.

  • C2PA provenance metadata is the future of platform-side auto-detection. Some AI music tools have signaled support for the standard. Until provenance is consistent end to end (and surviving re-encoding by video tools and platforms), the manual flag is what carries.
  • Meta's labeling policy applies to AI generated audio in Reels even when the visuals are not AI generated. The audio source is enough to trigger the rule.

4Do not rely on auto-detection for music

All three platforms have automated systems that attempt to detect AI generated content. Those systems are mostly trained on visual deepfakes (faces, scenes) and are less effective on music. A Suno or Udio track uploaded without the manual flag may pass the auto-detection check and still be subject to enforcement later if a human review concludes the disclosure was required. The safe pattern is manual disclosure on every AI music upload, regardless of what the auto-detection does. The cost of setting the flag is one toggle. The cost of missing it ranges from a warning to a monetization strike.

  • Provenance metadata (C2PA) is being adopted by AI tools and platforms. When the chain is intact, automatic labeling becomes more reliable. Today the chain breaks at many points (re-encoding, format conversion, screen capture), so manual flagging stays the default.
  • Tools that automate the metadata fill for you (like Dayvid's publish flow) still leave the AI disclosure as a deliberate per-upload choice, because that is what the platforms require.

5Keep records of what you flagged and why

For a monetized channel publishing AI music, keep a simple log of disclosure decisions. Date, platform, video title, flag set or not, reason. This is overkill for a hobby account, and standard practice for a channel that intends to grow into the Partner Program. If a platform ever asks about an upload (during a review, after a notice), the log is your evidence that you applied disclosure deliberately. Most creators never need this log, the ones who do need it badly.

  • A spreadsheet with one row per upload is enough. Channels using a tool like Dayvid can also rely on the project history as a partial log.
  • Treat AI disclosure logs as you would treat tax records. Boring until they matter.

6Stay current with each platform's policy text

AI disclosure policies updated multiple times in 2024 and 2025. The text on each platform's help center is the authoritative source, and it has changed before. Bookmark the three policy pages and re-read them every few months, especially before launching a new channel or a new monetization stream. Older creator guides will reference rules that have since been refined. This article links the current pages in the sources section; check them.

  • Policy changes often come with grace periods. If a rule changes, you do not need to retroactively re-label every upload, but new uploads under the new rule do need the new flag.
  • Platform help centers are searchable. "AI generated content" plus the platform name returns the policy page on each. Bookmark them and re-check.

Frequently asked questions

Creator reports across YouTube, TikTok, and Instagram suggest the disclosure label does not visibly suppress reach for AI music content in the typical case. The reach impact is far smaller than the consequences of failing to disclose when required. Set the flag, the cost is one click, the upside is no enforcement risk.

Most AI music with vocals from Suno or Udio is considered realistic for the purposes of YouTube's rule, because the model produces audio indistinguishable from a human-recorded performance for the average listener. Pure instrumental AI music falls under the rule as well in YouTube's current reading. When in doubt, disclose. The rule covers realistic audio that could be mistaken for non-AI content.

YouTube allows editing a video's content details after upload, including the altered content toggle, from YouTube Studio. TikTok and Instagram allow some post-upload edits depending on the post type. If you have a backlog of AI music uploads where you missed the disclosure, going back and setting it is the right move. It does not retroactively guarantee no enforcement, but it brings the channel into compliance and is what reviewers look for during YPP audits.

Yes, if your final upload contains AI generated audio, regardless of whether you generated it. The disclosure is about the audio your viewers hear, not about who originally produced it. Reaction, remix, and reupload content that contains AI music falls under the same rule. The original creator's disclosure on their version does not flow through to yours.

Dayvid publishes directly to YouTube as a private draft with the title, description, tags, and thumbnail filled in. The AI disclosure flag is a per-upload choice the creator makes on the publish step, because the platforms require the disclosure to be deliberate. Dayvid will not silently set or unset the flag for you. For TikTok and Instagram, Dayvid does not publish directly today; you download the rendered MP4 and upload natively, setting the AI label on each platform yourself.

Both platforms' current rules cover AI generated audio without a clear exemption for instrumentals or for background music. The conservative reading is that an AI generated track, whether vocal or instrumental, foreground or background, falls under the disclosure rule when it is realistic. Some creators interpret very short, clearly stylized clips (chip music, retro 8-bit) as outside the rule, but the safer call is to disclose anything AI generated until the platforms publish a more specific exemption.

Eventually, partially. C2PA is a content provenance standard that lets AI tools embed metadata in their outputs to identify what was AI generated. Suno, Udio, and major platforms have publicly engaged with the standard. When the chain is intact from generation through render to upload, platforms could auto-label correctly. The chain breaks today at many points (re-encoding, screen recording, format conversion in video tools), so manual disclosure stays the working default. Track this as it evolves.

Ready to make videos people watch?

Start free, no credit card. Generate your first video in under five minutes.

Get the rest of the publish flow rightSee plans and pricing

Related guides

Sources and methodology

Stats, figures, and external references cited in this guide were taken from the linked sources on the dates listed below. Information may be out of date by the time you read this.