Meta's "AI info" labels on ads: what changed on June 1, 2026 and what an ad label does not tell you
Meta's ads post, updated on June 1, 2026, says Meta is beginning to roll out "About this ad" and will start detecting ads made or edited with third-party AI tools. A label reports what Meta applied or detected. It does not tell you how the picture was made, or how much AI was used.
What changed on June 1, 2026?
Meta's newsroom post "Expanding GenAI Transparency for Meta's Ads Products" was first published on February 3, 2025. The page carries an update note dated June 1, 2026, which says the update reflects changes to the product. It does not list each change, so what follows is what the page says as of October 30, 2026.
- A new place for ad information. Meta says it is beginning to roll out "About this ad", a destination inside the three-dot menu on every ad.
- AI info labels in that place. Labels for ads made or significantly edited with Meta's own generative AI creative tools will appear there.
- Third-party tools. Meta says it will begin automatically detecting ads created or edited with third-party AI tools "through industry-standard signals", and will apply an "AI info" label in About this ad when it does.
The page gives no start date for the detection. It also says the experience may vary in some regions because of legal requirements.
What does Meta's ads page say, side by side?
| Topic | What Meta says |
|---|---|
| Ads made with Meta's tools | Images and videos made or significantly edited with Meta's generative AI creative tools get an "AI info" label. |
| Where the label sits | In the February 2025 text: in the three-dot menu or next to "Sponsored". When a tool produces a photorealistic human, next to "Sponsored". The June 2026 update says the label is included in About this ad. The page does not say whether the placement rules changed. |
| Ads made with other companies' tools | Meta will begin detecting them through industry-standard signals. The page does not name the signals or the tools. |
| What counts as a significant edit | The page does not define it. |
| Advertiser disclosure | The ads page does not describe a duty for advertisers to disclose AI use. Meta's April 2024 post on labeling separately says advertisers must disclose in certain cases when they digitally create or alter a political or social issue ad. |
Which signals does Meta say it reads?
The ads page says only "industry-standard signals" and does not name them. For posts, Meta's April 2024 post says it labels content when it detects "industry standard AI image indicators", or when people disclose that they are uploading AI-generated content. It says other companies put those indicators into content made with their tools.
Meta's February 2024 post names the standards for images from other companies: the C2PA and IPTC technical standards. That post is about posts, not ads, and the ads page does not say the same signals apply.
The same February 2024 post states the limits. It says people can strip out invisible markers, and that Meta could not identify all AI-generated content at the time. Meta's transparency page reports figures for October 1 to 29, 2024, mostly for organic content, and says the data may not capture labels on some content edited with AI. It does not mention ads.
What does an ad label not tell you?
These points are ours, drawn from what Meta's pages say and leave out.
- It is not a verdict on the pixels. Meta describes a label as the result of a detected signal, a tool Meta runs, or a disclosure. None of the pages we read calls the label proof that a picture is true or false.
- It does not measure how much AI was used. A "significant" edit is not defined, and a July 2024 update to Meta's April 2024 post says labels on content with minor AI modifications, such as retouching, were not always aligned with people's expectations (about posts).
- No label does not mean no AI. The ad may carry no signal Meta reads, or the signal may have been lost along the way.
- It says little about the advertiser's claim. A label reports AI use, not whether the product, price or offer shown is real.
How many AI images carry a signal like these?
The ads page does not name Meta's signals, so we cannot count what Meta reads for ads. We can count how many files carry the two kinds of signal our checker reads that fit the standards Meta names for posts: a signed Content Credentials record, or an unsigned IPTC declaration of how the image was made.
Measured October 30, 2026: of 31 real files filed as AI-made, 9 carry such a signal. That is 8 signed Content Credentials records and 1 unsigned IPTC declaration. Of the 8 records, 1 reads as trusted on the current C2PA Trust List, 5 match only the frozen Interim Trust List, and 2 read "signer trust not established". The other 22 files carry no such signal: 16 show nothing our checker reads, and 6 carry other metadata, such as a tool name or dates.
So 9 of 31, about 29%, carry a signal that could feed a detection built on signals like these. The files come from Wikimedia Commons categories named "Images generated by ..." and from two vendor samples. Third parties uploaded the Commons files, so we cannot know how each was handled. We ran no ad through Meta, so this is not a count of Meta labels. The tables are on our accuracy page.
What this means when you check an image
Our checker reads the file you give it. A screenshot of an ad is a new file and may carry none of the original's signals. Open the checker, choose the file, and read the headline. Two matter here.
- "No supported provenance signals found". The line under it says this does not mean the image is real or human-made, and that most images lose these signals when shared, screenshotted or re-saved. If an ad shows an "AI info" label, Meta may have used a signal your copy no longer holds.
- "Content Credentials found; signer trust not established". A signed record is present, but trust in its signer is not established, for example because the certificate is on neither trust list we load or has expired. The name is not verified, because anyone can make a certificate with any name. Meta's page does not say how it treats signer trust.
Our results say these checks cannot determine whether an image is AI-generated. They do not tell you whether Meta labeled the ad.
Guides on this topic
- Visible AI watermarks is the main page on this topic.
- Meta's "AI info" label covers organic posts on Facebook, Instagram and Threads.
- Meta's invisible watermark covers what Meta says it adds to its own images.
- Which platforms label AI images compares the labels.
- YouTube's altered or synthetic content label covers a platform that asks creators to disclose.
Frequently asked questions
- Where do I find the AI info label on a Meta ad?
- Meta says it appears in About this ad, in the three-dot menu on an ad, and for some ads next to the "Sponsored" label. Meta says the experience may vary by region.
- Does Meta label ads made with other companies' AI tools?
- Meta says it will begin automatically detecting them through industry-standard signals. The page names neither the signals nor the tools, and gives no date.
- Does an AI info label mean the ad is misleading?
- Meta's pages we read do not say so. A label reports a detection or a tool, not whether the ad is true.
- Can an ad made with AI show no label?
- Meta's February 2024 post says people can strip out invisible markers and that Meta could not identify all AI-generated content at the time. A missing label is not evidence that a picture is real or human-made.