Start with evidence, not assumptions
Image metadata is often spread across EXIF, IPTC, XMP, and the file container itself. A team can think a photo is clean because it looks ordinary, while the file still reveals capture location, device details, author names, copyright text, or editorial notes. The first job is to inspect the whole footprint before changing anything.
Fix orientation before deciding on cleanup
Many image files are not stored in the direction they appear on screen. They rely on an orientation tag. If you skip that step, you can end up exporting a privacy-clean file that is still sideways for the recipient. Normalize orientation first so every later review is based on the real delivered appearance.
Decide whether the file should stay descriptive or become anonymous
Some workflows need metadata to remain useful. Newsrooms, archives, and marketing teams may need copyright, caption, creator, or usage information to survive handoff. In that case, editing the remaining fields is better than deleting everything blindly. Other workflows prioritize privacy and safe sharing. For those, the correct outcome is an irreversible cleaned export that removes location, device, and identity clues.
Verify the exact exported format
Privacy cleanup is only complete when the final JPEG, PNG, or WebP export has been reopened and checked again. That confirmation should show the image is upright, the required visible content is intact, and the unwanted metadata no longer appears anywhere a downstream recipient can read it.