Which platforms read or show Content Credentials? LinkedIn, TikTok, YouTube, Meta and Google
Five large companies describe Content Credentials support, and each describes something different. LinkedIn says it shows a C2PA icon, TikTok says it reads the records, YouTube says it may apply a label, Meta says it looks for industry-standard signals, and Google says its checks are arriving. We did not test any platform ourselves.
What does each platform say about Content Credentials?
Platform features change often, so check each source before relying on a row.
| Company and source | What the page says | What the page does not say |
|---|---|---|
| LinkedIn help page | Image and video content signed with C2PA Content Credentials is noted with a C2PA icon. Clicking it shows the credential and available metadata, such as assertions about AI use, the app or device, the creator and the issuer. It says the icon is visible to all viewers of your content. LinkedIn says it is gradually making the feature available and you might not have access. For sponsored content, it says the icon appears on single image or video ads only when the image or video is uploaded. | That LinkedIn checks whether a signer is trusted, or what it does with files that carry no record. The page shows only a relative "last updated" note, so we cite no date for it. |
| TikTok newsroom post, November 19, 2025 | TikTok says it reads C2PA Content Credentials and adds them to AI-generated content made on TikTok. It says it was testing "invisible watermarks" and would start adding them to content made with TikTok tools and to content uploaded with Content Credentials. It says labels may be removed when content is reuploaded or edited elsewhere. | Whether the watermarks have started, or how TikTok decides what to label. |
| YouTube help page | YouTube says it may automatically apply an AI label to content made with its own GenAI tools, to content that contains C2PA metadata, and to content its internal systems detect as generated or altered. It says creators cannot adjust the label on content made with its AI tools, content containing C2PA metadata or content labeled after manual review. | How YouTube reads the metadata, or which signers it accepts. |
| Meta post on Facebook, Instagram and Threads, dated February 6, 2024 (the page also shows April 1, 2025) | Meta says it will label images that users post to Facebook, Instagram and Threads when it can detect industry-standard indicators that they are AI-generated. It says it is building tools that can identify invisible markers at scale, including the "AI generated" information in the C2PA and IPTC standards. It says photorealistic images made with its Meta AI feature carry invisible watermarks and metadata. It says there are ways people can strip out invisible markers and that it is not yet possible to identify all AI-generated content. | Which signers Meta accepts, or what happens to a record after upload. |
| Google blog post, May 19, 2026 | Google says a Content Credentials check, for whether media is an unaltered camera original or has been modified, was rolling out in the Gemini app that day and would come to Search and Chrome "in the coming months". The same post says Meta will start labeling camera-captured media with Content Credentials on Instagram. | Whether Search and Chrome have received the check. The page shows no update notice. The Instagram sentence is Google's statement; we read no Meta page that says it. |
Does Instagram support C2PA?
Meta's 2024 post, which covers Facebook, Instagram and Threads, says Meta is building tools that can identify invisible markers, including the "AI generated" information in the C2PA and IPTC standards, and will label images when it can detect industry-standard indicators. It does not say how Meta treats a signed record after upload.
Which platforms appear on the trust lists?
A trust list names signers, not the platforms that read records. Still, we read the subject of every certificate in the three files that hold signers, to see which of these names appear. Our copies were fetched October 27, 2026, and the current C2PA Trust List copy is dated October 17, 2026. The files hold 32, 27 and 115 certificates.
| Name in the certificate | C2PA Trust List anchors | Interim anchors | Interim allowed list |
|---|---|---|---|
| Pinterest Inc. | 0 | 1 | 0 |
| TikTok Inc. | 0 | 0 | 1 |
| Microsoft Corporation | 0 | 5 | 0 |
| Adobe | 1 | 1 | 0 |
| Google LLC | 6 | 0 | 0 |
| Amazon Web Services | 0 | 0 | 15 |
| Cloudflare | 0 | 0 | 1 |
None of the 174 certificates names LinkedIn, Facebook, Meta, Instagram, YouTube, X, Snap, Reddit or Canva. A name here is what the certificate says, not proof that a company signs what it makes. Our guide to the two lists explains them.
What happens to a record when a picture is re-rendered?
We have no test of any platform's upload. We do have one related measurement. On October 30, 2026 we re-rendered the 9 real files in our set that carry a Content Credentials record. We drew each one on a browser canvas and saved it again as a PNG and as a JPEG, which is what a screenshot or a re-encode does to the pixels. After the PNG re-render, 0 of 9 still carried a record. After the JPEG re-render, 0 of 9 did. All 18 outputs read "No supported provenance signals found".
This was one browser, one JPEG quality and a canvas, not a platform. It shows that a record is lost when pixels are re-drawn, and nothing about what any platform does. Our guide to why Content Credentials disappear reports the same result, and our accuracy page shows how we test the checker.
Is a platform label the same as a trust check?
No. The LinkedIn page we read describes the icon as opening the record's contents, such as the issuer and the date the credential was issued. It does not say the signer was checked against a trust list. A label on its own does not tell you whether the certificate sits on a list you trust, and a signed record shows who signed and that the bytes match, not that the claim is true.
What this means when you check an image
Open the checker, choose the original file you downloaded, and look for the headline. Five of the 9 real files we hold with a record read "Content Credentials found; legacy trust-list match". That headline means a signed record is present and the signer is on the frozen Interim Trust List, not the current C2PA Trust List. It says nothing about what any platform showed.
If the file went through a screenshot or a re-save, you may instead see "No supported provenance signals found". That does not mean the image is real or human-made. The checker's limits line applies: these checks cannot determine whether an image is AI-generated. Our Privacy and method page explains what happens to your file.
Guides on this topic
- What are Content Credentials? is the main page on this topic.
- What the LinkedIn Content Credentials label means covers one platform in depth.
- YouTube's altered or synthetic content label explains the disclosure rules.
- Which platforms label AI images? compares labels beyond Content Credentials.
Frequently asked questions
- Does LinkedIn show Content Credentials?
- LinkedIn's help page says content signed with C2PA Content Credentials is noted with a C2PA icon that opens the credential, and that the feature is being made available gradually. The page does not say LinkedIn judges whether the signer is trusted.
- Does TikTok read Content Credentials?
- TikTok's November 19, 2025 post says it reads C2PA Content Credentials and adds them to AI-generated content made on TikTok. It adds that labels may be removed when content is reuploaded or edited elsewhere.
- Will a platform keep the record on my image?
- We did not test that. Our canvas re-render left a record in 0 of 9 files, but that was not a platform. Check what the platform says, and keep the original file.
- Is a platform showing no label proof that a record is missing?
- No. A platform may not have rolled the feature out to you, or may not have read the file.