isGenAI.com

AI image detector by platform

Threads AI image detector: Meta’s label, Meta’s re-encode

Threads follows Meta’s policy of labelling images with “AI info” when C2PA or IPTC AI indicators are read at upload, and serves re-encoded copies that normally carry none of those records, so the Threads label is a platform observation and the original file is what isGenAI needs.

Platform behaviour as documented or widely reported in October 2026. Updated .

Check the downloaded imageFrequently asked questions
Code-drawn illustration for Threads

Illustration · PNG file

  1. Signature + IHDR
  2. iTXt XML:com.adobe.xmprecord
  3. tEXt parameters
  4. tEXt prompt / workflow
  5. tEXt invokeai_metadata
  6. eXIf
  7. caBX C2PA manifest
  8. IDAT pixels
  9. IEND

AI source type declared

The file labels itself as AI-made.

IPTC DigitalSourceType such as trainedAlgorithmicMedia is read from XMP and shown with its exact value.

Check an image →
AI source type declared: where this record sits inside a PNG file and what the isGenAI report says about it. The picture is a code-drawn illustration.

What Threads strips and adds

Meta’s announcement covers Facebook, Instagram and Threads together: industry-standard AI indicators in C2PA and IPTC metadata are read at upload to label posts, Meta AI images carry visible markers, invisible watermarks and metadata, and Meta states that not all AI content can be identified and that markers can be stripped. Threads images are resized and re-encoded for delivery and in practice arrive without EXIF, XMP or a manifest.

Cross-posts from Instagram to Threads pass through the same pipeline.

What isGenAI shows for a Threads download

The AI image detector reads the file you upload: a C2PA manifest if one survived, the IPTC DigitalSourceType in XMP, AI statements in metadata, saved prompts in the AUTOMATIC1111, ComfyUI and InvokeAI formats, and matching copies by hash and perceptual fingerprint. A platform copy with none of these returns “no evidence found”, and the report says plainly that absence proves nothing about the author.

Record the label separately as a platform observation; it answers what Meta read at upload, not what the file you hold contains.

What to do instead

Ask the poster for the original export, or trace the earliest appearance of the image. The matching-copy check can connect a stripped Threads copy to an earlier file in isGenAI’s records.

The Instagram and Facebook pages describe the same Meta pipeline.

How to check an image from this platform

  1. Note the platform’s own label or credential icon on the post before you download anything; it does not travel with the file.
  2. Ask the poster for the original file, or trace the first upload. A platform download is a re-encoded copy on most services.
  3. Upload what you have to the AI image detector and read the report. A stripped copy returns “no evidence found” honestly.
  4. Use the matching-copy check: a hash or perceptual fingerprint match can tie the stripped copy to an original that still carried records.

Check the downloaded image

Frequently asked questions

Does Threads strip image metadata?

Threads serves re-encoded copies that arrive without EXIF, XMP or C2PA records in practice. Meta documents reading those records at upload to label posts.

What is the Threads “AI info” label based on?

C2PA or IPTC AI indicators read in the upload, Meta AI’s own markers, or the poster’s disclosure, per Meta’s announcement.

Can isGenAI detect AI in an image saved from Threads?

Only from surviving records or a matching copy. Most Threads downloads return no evidence found; check the original file.

Does no label mean the image is a real photo?

No. Meta states it cannot identify all AI-generated content. Trace the source and check the original.

Sources

Where a platform documents its behaviour, the vendor page is linked. Where it does not, the page says the behaviour is reported, not documented.