Facebook, Instagram and Threads: AI content rules and Content Credentials
Short answer
Meta requires people to use its disclosure tool when they post organic content with photorealistic video or realistic-sounding audio that was digitally created or altered, and adds “AI info” labels when it detects industry-standard AI signals, including the C2PA and IPTC information, or when people disclose. Advertisers must disclose AI-made or AI-edited photorealistic images, video or realistic audio in ads about social issues, elections or politics.
Official sources: Meta: Labeling AI-Generated Images on Facebook, Instagram and Threads · Meta: Our Approach to Labeling AI-Generated Content and Manipulated Media
Covers: Facebook · Instagram · Threads
| Label required from creators | Yes, photorealistic video and realistic audio; political ads |
|---|---|
| Platform labels by itself | Yes: C2PA and IPTC signals |
| Content Credentials (C2PA) | Reads C2PA and IPTC information |
| Read on | 3 November 2026 |
AI content rules: what Facebook, Instagram and Threads says
Disclosure required
Rule: discloseMeta’s February 2024 announcement.
“We’ll require people to use this disclosure and label tool when they post organic content with a photorealistic video or realistic-sounding audio that was digitally created or altered, and we may apply penalties if they fail to do so.”
Edited content: label in the menu
Platform labelMeta’s update of 12 September 2024.
“For content that we detect was only modified or edited by AI tools, we are moving the “AI info” label to the post’s menu.”
A more prominent label
Platform labelMeta’s April 2024 post.
“If we determine that digitally-created or altered images, video or audio create a particularly high risk of materially deceiving the public on a matter of importance, we may add a more prominent label so people have more information and context.”
Political and social issue ads
AdsMeta’s ad policy page, updated 19 February 2026. The listed cases: a real person saying or doing something they didn’t; a realistic-looking non-existent person; a realistic event that didn’t happen; altered footage of a real event; a realistic, alleged event that isn’t a true recording. Immaterial uses such as resizing or colour correction need no disclosure.
“Advertisers must still disclose when ads about social issues, elections or politics use AI to create or edit photorealistic images, videos or realistic audio that depicts: A real person saying or doing something that they didn't.”
If advertisers do not disclose
EnforcementSame page.
“Failure to disclose the scenarios above may result in ad removal and account penalties for repeated violations.”
Ad label renamed
AdsSame page.
“The social issues, elections and politics (SIEP) self-disclosure label will soon change from "digitally created" to "AI info".”
If you post AI content on Facebook, Instagram and Threads
- Use Meta’s AI disclosure when you post photorealistic video or realistic-sounding audio made or altered with AI.
- Keep the C2PA or IPTC metadata from your AI tool: Meta reads it to add “AI info” labels.
- For ads about social issues, elections or politics, tick the AI disclosure in the ad flow, or the ad may be rejected.
- Under the EU AI Act, a professional deepfake needs your own clear disclosure too; Meta’s label is the platform’s.
Content Credentials (C2PA): what the source says
Reads C2PA and IPTC information
Documented: reads and labelsMeta’s February 2024 announcement covering Facebook, Instagram and Threads together. It is forward-looking: it describes tools being built.
“We’re building industry-leading tools that can identify invisible markers at scale – specifically, the “AI generated” information in the C2PA and IPTC technical standards”
“AI info” labels
Documented: reads and labelsMeta’s April 2024 policy post (updated 1 July 2024). This sentence says “industry standard AI image indicators” and does not name C2PA or IPTC.
“We will begin adding “AI info” labels to a wider range of video, audio and image content when we detect industry standard AI image indicators or when people disclose that they’re uploading AI-generated content.”
What Meta says about removal
Documented: removalFrom the February 2024 post: markers can be stripped by people. This is a statement about invisible markers, not about what the apps do to uploaded files.
“there are ways that people can strip out invisible markers”
What is and is not documented
- The statements apply to the three apps together; Meta’s posts do not set out per-app behaviour.
- The label was renamed from “Made with AI” to “AI info” on 1 July 2024, and for content only edited by AI the label moves to the post menu, according to the April post.
- We found nothing in Meta’s pages we could read about whether the apps keep or remove the metadata inside an uploaded file. Meta’s help-centre pages were not retrievable.
Check a file yourself
The Content Credentials checker reads the C2PA data in an image in your browser and shows who signed it, the generating tool and whether the signature still matches. It reads what a file declares; it does not detect AI from pixels, and a file with no credentials proves nothing either way. See why metadata gets stripped.
Questions
- Do I have to label AI content on Facebook, Instagram and Threads?
- Facebook, Instagram and Threads’s page says: “We’ll require people to use this disclosure and label tool when they post organic content with a photorealistic video or realistic-sounding audio that was digitally created or altered, and we may apply penalties if they fail to do so.” (Labeling AI-Generated Images on Facebook, Instagram and Threads).
- Does Facebook, Instagram and Threads keep or strip Content Credentials?
- Documented: Meta says it is building tools to read “AI generated” information in the C2PA and IPTC standards, and that it adds “AI info” labels when it detects industry-standard AI image indicators. Meta also says people can strip out invisible markers.
- Does Facebook, Instagram and Threads read Content Credentials to label content?
- The sources we read say so: “We’re building industry-leading tools that can identify invisible markers at scale – specifically, the “AI generated” information in the C2PA and IPTC technical standards” (Labeling AI-Generated Images on Facebook, Instagram and Threads).
- What is not documented about Facebook, Instagram and Threads?
- The statements apply to the three apps together; Meta’s posts do not set out per-app behaviour. The label was renamed from “Made with AI” to “AI info” on 1 July 2024, and for content only edited by AI the label moves to the post menu, according to the April post. We found nothing in Meta’s pages we could read about whether the apps keep or remove the metadata inside an uploaded file. Meta’s help-centre pages were not retrievable.
This page reports what the cited sources say on the date shown. It is not an assessment of anyone’s product, it is not legal advice, and Proofcan does not guarantee compliance. Spotted something out of date? Write to support@proofcan.com.
Sources
- Labeling AI-Generated Images on Facebook, Instagram and ThreadsMetalast checked
- Our Approach to Labeling AI-Generated Content and Manipulated MediaMetalast checked
- Disclosure of digitally created or altered content in political or social issue ads (updated 19 February 2026)Metalast checked
Page reviewed . Proofcan is a technical tool, not legal advice, and does not guarantee compliance.