"AI info" label

How to read it

Meta says it adds the label when it detects industry-standard AI indicators in a file, or when people disclose that they are uploading AI-generated content (as Meta’s newsroom and Transparency Center describe it, September 2026). The indicators are the “AI generated” information in two metadata standards, C2PA and IPTC, which Meta said in February 2024 it would read in images from Google, OpenAI, Microsoft, Adobe, Midjourney and Shutterstock. Meta started labelling in May 2024 with the words “Made with AI” and changed them to “AI info” on 1 July 2024. Images made with Meta’s own AI feature have carried the label “Imagined with AI”.

Also searched as: AI info Facebook, AI info Instagram.

Why “Made with AI” became “AI info”

Photographers complained, TechCrunch reported, that Meta was putting “Made with AI” on real photos that had only been through some basic editing tools. Meta agreed that its labels sometimes didn’t match what people expected: some content with minor AI edits, such as retouching, carried the industry-standard indicators and was labelled. Meta changed the wording but not the detection, which still reads C2PA and IPTC metadata, so TechCrunch noted that a photo with an object removed by Adobe’s Generative Fill might still be labelled. IPTC’s own vocabulary separates media “Created using Generative AI” from media “Edited using Generative AI”, such as inpainting and outpainting.

Where the label appears

From September 2024, Meta moved the label into the post’s menu for content it detects was only modified or edited with AI tools. For content it detects was generated by AI, the label stays on the post, and Meta says whether it is there because of industry signals or because someone self-disclosed. For content that risks materially deceiving the public on a matter of importance, Meta says it may add a more prominent label.

What it is based on

Meta’s label depends on signals other companies’ tools put into files and on people disclosing. Meta requires disclosure for a photorealistic video or realistic-sounding audio that was digitally created or altered, and may apply penalties to people who don’t. It says it isn’t yet possible to identify all AI-generated content and that people can strip out invisible markers, so a post can be AI-generated without carrying the label. Its Transparency Center calls its labelling method still evolving. Over 29 days in October 2024, Meta counted over 1 trillion views of the AI info label on Instagram.

Look-alikes

“Made with AI” is the label’s old name, and other platforms use similar words ("Made with AI" label). The CR pin of Content Credentials (CR) icon reads the same C2PA data but is a separate icon. Google Photos also has an AI info section, in a photo’s Info panel; Google removes it for photos that haven’t been edited with AI or don’t follow the IPTC standards. The Instagram icons page covers Instagram’s other icons.