Does Social Media Remove AI Image Metadata?

Platforms often resize, recompress, transcode, or sanitize uploaded images. These steps can remove EXIF, XMP, C2PA manifest stores, or provider-specific marker data. Some removal is intentional privacy protection; some is a side effect of media processing.

A screenshot is even more destructive for provenance because it creates a new file from rendered pixels. The screenshot may preserve what the image looks like, but it usually does not preserve the original file's creation history.

Upload the original file from the creator, camera, generator, or editing tool whenever possible. If you only have a social download, label the result as a check of that downloaded copy, not the original asset.

If the image may affect safety, identity, elections, health, or reputation, ask the source for the original file before making a strong claim.

No C2PA, no EXIF, or no AI marker found means the specific file did not expose those signals to the checker. It does not prove the image is real, fake, AI-generated, or human-shot.

Use wording like no verified provenance found or metadata not present in this copy. That language is slower, but it is much harder to misuse.

Example: a user uploads an image saved from a social platform. Img2det finds no C2PA and weak metadata, but the user later uploads the original file and receives marker or credential evidence. The difference does not mean the first report failed. It means the platform copy likely lost the evidence needed for a stronger conclusion.

This is the kind of result where a low-value detector page would usually stop too early. Img2det keeps the finding tied to the evidence category that produced it, then separates strong provenance, marker-only clues, camera-like support, and uncalibrated forensic context. That separation gives a reviewer something they can cite or challenge instead of a single unexplained score.

For social-media metadata loss, compare report fields across versions. Look for changes in C2PA presence, EXIF fields, byte marker context, file type, dimensions, and compression-related frequency notes. The guide should teach users to compare copies instead of judging one stripped file in isolation.

Read the Final assessment first, then open the detailed evidence matrix. The matrix shows whether the strongest available signal came from verified C2PA provenance, an AI-related marker, EXIF or camera-like metadata, byte-level context, or frequency analysis. If the top signal is weak, the right conclusion is usually uncertainty, not a stronger accusation.

Check the social download first if that is all you have, but label the result as platform-copy evidence. Then request the original upload, a creator export, or an archived source. Run each copy and record which signals survived. This turns metadata loss into an evidence trail rather than a dead end.

When the result will be used for editorial, moderation, or public claims, save the report language exactly as evidence language. Use phrases such as marker found, no verified manifest, camera-like support, or inconclusive. Avoid rewriting those into definitive claims like fake, real, generated, or authentic unless you also have source context outside the file.

The false positive is saying no metadata means fake or AI. The false negative is saying no marker means not AI. Social platforms commonly recompress files, resize images, strip EXIF, and remove embedded provenance during processing.

The opposite error is also common: treating a quiet report as proof that the image is camera-original. A quiet report may simply mean that useful metadata was removed, that the generator did not add supported credentials, or that the file was exported through a workflow that stripped the strongest signals.

People return because platform behavior is part of the investigation. A guide that explains why a social copy is weaker helps users come back with the right file and understand why the second report is more informative.

That repeatable review flow is the practical value of the site. Users can run an image, compare the result with the guide language, and return when they receive a better original file or a new version from another platform. The content supports the tool instead of acting as a doorway page for one keyword.