How to Flag AI Content on Instagram & Facebook
When you must disclose
- ⚑ Photorealistic AI-generated images that could be mistaken for a real photograph.
- ⚑ AI-generated or AI-manipulated video and audio that depicts realistic people, places or events.
- ⚑ AI-generated content in ads — Meta Ads Manager has a disclosure control advertisers are required to use for AI creative.
- ⚑ Realistic synthetic media added in editors that do not embed industry-standard metadata (for example some composites).
When you usually don't
- ✓ Minor AI retouching and adjustment tools that do not change what the content depicts.
- ✓ AI used purely for productivity around the post, not in the media itself.
- ✓ Obviously unrealistic or stylised images no viewer would take as a real photo.
- ✓ Content where a human has made only trivial, non-deceptive edits.
When in doubt, disclose — platforms can apply a label themselves and penalise repeated non-disclosure.
How to disclose AI content on Instagram & Facebook
- When creating a post or Reel on Instagram or Facebook, open the advanced settings before sharing.
- Find the option to indicate the post contains AI-generated or digitally created content and switch it on.
- Meta adds an “AI Info” label, which viewers can tap for more detail about how the content was made.
- If your media carries C2PA Content Credentials from the AI tool, Meta’s systems may detect and label it automatically.
- For ads, use the disclosure control in Meta Ads Manager when your creative contains AI-generated or AI-manipulated content.
- Add your own visible caption disclosure and embed metadata in AI images so the label is portable and reliable (see the generator below).
Generate your disclosure label & metadata
Instagram & Facebook's in-app toggle marks the upload itself. This free generator gives you the extra layer: a visible badge for your caption or description, a plain-text disclosure, and machine-readable metadata embedded into your image file — which platforms also read automatically.
Create your disclosure
Your disclosure
Fill in the form and your badge, metadata and plain-text disclosure will appear here.
How Meta's "AI Info" label works
Meta applies a label — now shown as “AI Info” — to video, audio and image content across Instagram, Facebook and Threads. The label appears in two situations: when Meta detects industry-standard signals that content is AI-generated, and when a creator self-discloses during posting. Tapping the label shows viewers whether it was applied because of detected signals or because someone disclosed it.
Meta originally used a “Made with AI” label but changed it to “AI Info” after feedback that the older wording was applied too broadly — for example to photos that had only minor AI retouching. The current label is designed to give context rather than imply a photo is entirely fake.
Detection through C2PA Content Credentials
Meta’s automatic detection relies heavily on C2PA Content Credentials — verifiable provenance metadata attached by AI tools such as Adobe Firefly and others. When you upload an image carrying that metadata, Meta reads the embedded manifest (which can include the creation tool and timestamp) and adds the label without any action from you.
This has a practical catch: false positives happen. Legitimate product or studio photography has sometimes been labelled because the file retained metadata from an earlier edit in an AI-capable editor. And detection fails when metadata is stripped. So manual self-disclosure remains important — both to catch what detection misses and to stay in control of how your content is described.
Advertisers: Meta's rules are enforced
For advertisers, AI disclosure on Meta is not optional. Meta Ads Manager includes a disclosure control that advertisers are required to use when ad creative contains AI-generated or AI-manipulated content, and Meta’s automated systems also apply the label on detection. The label is not cosmetic: where AI content is judged deceptive — particularly realistic media depicting events that did not occur — Meta can apply a more prominent label and substantially reduce distribution.
Meta can also retroactively flag a campaign that was approved while running, as its classifiers expand from images into audio and video. The safe posture is to disclose before launch rather than correct afterwards. Bear in mind too that the EU AI Act and various US state laws impose their own disclosure duties that can exceed Meta’s platform rule — meeting Meta’s label alone does not guarantee legal compliance.
This page is general information, not legal advice, and is not affiliated with or endorsed by Instagram & Facebook. Platform policies change; verify against the platform's official help pages.
Instagram & Facebook AI disclosure — FAQs
How do I flag a post as AI-generated on Instagram?
When creating the post or Reel, open the advanced settings before sharing and turn on the option indicating the content is AI-generated or digitally created. Meta then adds an "AI Info" label.
What is the "AI Info" label on Facebook and Instagram?
It is the label Meta applies to AI-generated video, audio and images. It appears when Meta detects industry-standard AI signals or when a creator self-discloses. Viewers can tap it for more information.
Why did Meta label my real photo as AI?
Meta's detection reads embedded metadata. If your photo was exported from an AI-capable editor, it may retain metadata that triggers the label even though the photo is genuine — a known false-positive issue. You can check the file's metadata before posting.
Do I have to disclose AI in Meta ads?
Yes. Meta Ads Manager has a disclosure control advertisers are required to use when creative contains AI-generated or AI-manipulated content. Undisclosed AI ad creative can be rejected or flagged retroactively.
Does Meta detect AI content automatically?
Partly. Meta reads C2PA Content Credentials and uses detection classifiers. But metadata can be stripped and detection is imperfect, so self-disclosure is still expected.
Does the "AI Info" label reduce my reach?
The label itself is informational. However, AI content judged deceptive — realistic media of events that did not happen — can receive a more prominent label and significantly reduced distribution.
More disclosure tools & guides
Main disclosure generator → All guides
See also: YouTube · TikTok · Instagram & Facebook