AiMetaOff / GUIDES

X (Twitter) photo metadata, GPS and AI labels

See what X says about photo EXIF, GPS and image downloads. Check post location, understand Grok and AI labels, and inspect the file before uploading.

X says people viewing a photo on the platform do not get its EXIF data. But X can read EXIF from the file you submit, and a location attached to the post can still be visible. If you are posting a photo taken at home, check both the file you are about to attach and the location shown in the post composer. X Help on posting photos · X Help on post location

This guide draws on published documentation. Live X uploads and Grok export comparisons were not independently tested.

Does X (Twitter) remove EXIF and GPS from photos?

The original upload is the file you attach. It may contain EXIF capture details, EXIF GPS coordinates, XMP, IPTC or signed Content Credentials. With location services enabled, X says it can use photo EXIF to suggest places in the composer. X also describes using EXIF to analyze whether an image came from the device camera and keeping it temporarily for photo processing. The help page gives no specific retention period. X Help on posting photos

An X-served photo is the version X delivers for viewing. X says people viewing the photo do not have access to its EXIF, which includes coordinates stored in the EXIF GPS fields. This statement covers viewer access to EXIF in the documented photo-post workflow. The help page leaves other metadata formats, direct messages, profile images and the handling of each downloadable variant open. A clean-looking download can tell you about that file. It cannot show what X received at upload. X Help on posting photos

Embedded GPS and X post location

A photo can carry camera coordinates inside the file, and the post can separately carry a place label beneath its text. On x.com, X describes selecting a label such as a city or neighborhood. On iOS and Android, X says a post made with its in-app camera can include the chosen label and the device’s precise latitude and longitude if precise tagging is enabled for that post. X says those coordinates can be accessed through its API. A broad place label can therefore accompany more specific coordinates. X post location FAQs

Before posting, open the location picker and clear any selected place if you want to keep it private. Check even if the previous post was the only one you intended to tag. X says a general label can carry over to the next post. To limit device location access, review X’s precise-location setting and the app’s permissions on your phone. For older posts, X documents a bulk option for removing past post location information and warns that outside copies can persist. X posting location controls · X precise-location controls for mobile · X post location FAQs

What file do you get when you save an X photo?

Open the photo from its post before saving it, especially if you reached it through a timeline preview. X documents scaling for display, and its developer reference lists a photo URL and media dimensions. The reference gives no byte-for-byte guarantee for the upload and says little about how a particular client saves an image. Even a copy with the same pixel dimensions as the source can have different metadata. X Help on posting photos · X API media fields

If you are checking what other people can download, inspect the actual saved file and note which app or browser saved it. If you need to know what the uploader sent, ask for the source file. Keep the post link when you need to record a location label or AI notice alongside the download.

A native Repost references the existing post. Saving the photo and attaching it to a new post submits that saved file as a new upload. A screenshot makes a new image of what was visible on screen and can capture a location label or watermark as pixels. A conversion makes another file whose metadata depends on the exporter and its settings. Inspect the final file after either operation. X Help on Reposts and Quotes

What does X’s synthetic and manipulated media label mean?

Read the exact wording on the post. X describes two relevant labels, and they answer different questions.

Manipulated media is a notice X may apply when it determines that media was significantly and deceptively altered or fabricated. X’s authenticity rules address misleading media that can cause serious harm, affect public safety or cause widespread confusion on public issues. Read any context linked from the notice to see what X is flagging. If you believe a post violates X’s rules, use the post menu’s reporting option. X notices · X authenticity policy · Report a post
Made with AI is an indicator X uses for detected AI-generated media. X says its C2PA integration contributes to automatic labeling, and its Media Literacy Action Plan specifically describes expanded indicators for users in the EU. These statements leave the exact coverage by client, region and image workflow open. X C2PA integration · X Media Literacy Action Plan

A Community Note is context added by contributors. A provider watermark, such as Grok’s, is a mark on the media. If you see both a Note and a label, read each where it appears and record the wording if you need to report a problem. X Help on Community Notes · Grok website and app FAQ

Does X preserve C2PA Content Credentials?

X publicly describes using C2PA to help verify provenance and label detected AI-generated content. Those statements describe what X can use for detection. They leave open whether X keeps an embedded C2PA manifest in every image copy it serves or makes available to save. A copy with no readable manifest leaves the source upload and any signals X saw unresolved. X C2PA integration

A Content Credential is a signed record bound to a media asset. It can carry claims about capture, editing or AI generation. If the original post has a label and you have a copy of the file, read each on its own terms: the post shows what X displayed, and the file gives you the record your inspector could read. C2PA explainer · Understand what embedded credentials can establish

Grok images and AI provenance on X

Generating an image with Grok and attaching an image to an X post are separate steps. The Grok website and app FAQ describes a Grok watermark on generated images and videos. X also describes Grok watermarks in its India guidance. A watermark on the image, a signed credential inside a file and an X label next to a post each have different sources. Grok website and app FAQ · X guidance for India

X’s Grok help describes conversation history, inputs and results. The materials reviewed here give no general specification for which prompts, generation records or C2PA manifests are embedded in a Grok image download. If you have the downloaded file, inspect that copy and keep the provider’s provenance signals intact. Follow the disclosure instructions shown when you post. About Grok on X · Grok website and app FAQ

Inspect the photo you plan to attach

Pick the exact file you intend to attach in the X composer. If you cropped or exported it, use that last export.
Open it in AiMetaOff’s Metadata Viewer, called Inspect File on the site. It reads supported records locally in your browser.
Check EXIF GPS first. Also look for a place name or address in readable XMP, IPTC, captions or text fields. Review any unreadable or unsupported entries and keep any embedded provenance you need to share.

Inspect a copy of the original image

Remove supported private fields and check the downloaded copy

For a supported image, use the Image Metadata Remover workspace to review the cleanup selections. Select the categories that contain private information you want removed. EXIF cleanup works on the category as a whole, so it can remove capture details and thumbnails along with GPS. Minimal orientation may remain to keep the image displaying correctly. If the same location is stored in another supported category, review that category as well. Check the selections before cleaning so you can preserve credentials and other records you intend to keep.

Start cleaning and download the new copy. Open the download in Inspect File and confirm that the supported fields you targeted are absent. If the report marks a check as unsupported or the result as partial, resolve that before relying on the output. Open the photo itself too. A street number, shop sign or location printed by a camera app can still identify the place. Finally, attach the checked copy and clear the composer’s location if one appears. Guide to verifying metadata removal

Frequently asked questions

Can someone see where I took my photo on X?

A viewer may see a place you attached to the post, and precisely tagged in-app camera posts can expose coordinates through the API. People can also recognize landmarks in the photo. X’s statement about viewer access applies to EXIF embedded in the photo. X post location FAQs · X Help on posting photos

Does downloading a Twitter photo preserve EXIF?

X says viewers do not get the photo’s EXIF through its documented posting flow. If you need to know what any particular saved version contains, inspect that file. To check what was in the upload, get the source file from the uploader. X Help on posting photos

Can I find the Grok prompt in an image saved from X?

Official Grok documentation describes image output and the watermark, but gives no general promise that an image embeds its prompt. If the prompt matters, ask the creator for the generation record or a prompt they can share. About Grok on X · Grok website and app FAQ

Does Made with AI prove the downloaded image contains C2PA?

The label shows X’s assessment on the post. It does not identify which signals were used for that specific photo or certify the contents of the file you saved. X C2PA integration

Will removing GPS change an X AI label?

GPS cleanup addresses location fields in your file. X applies its own labels to posts and documents C2PA as one contribution to detection. Use the post notice to check what X displayed. X notices · X C2PA integration

Sources

Official sources checked on November 14, 2026.