What xAI says about marking
xAI's FAQ says generated images and videos carry a visible Grok watermark, and its usage policy treats removing provenance signals as a violation. Grok Imagine 1.5 went into wide release on 17 June 2026 with image-to-video generation, and the same marking applies.
Because xAI documents no C2PA manifest, images posted on X are not labeled automatically by systems that read Content Credentials.
Why check a Grok file anyway
Files pick up metadata as they move: an editing app may add XMP, a phone may add EXIF when you save to the gallery. BlankAI reports exactly what is present, so you know what you are sharing.