Add Copyright to Photo EXIF Data

EXIF is mostly a record of what the camera did — shutter speed, ISO, lens, timestamp. But the standard reserves two string fields for the human being behind the camera, and those are the ones you want:

That is the whole of what EXIF offers. There is no field for your email address, no field for licensing terms, no field for a link to your website. If you want those — and if you are selling work, you do — they live in the IPTC and XMP blocks, which sit in the same file and carry Creator, Copyright Notice, Rights Usage Terms, and a Web Statement URL.

This is the step most photographers skip, and it is the one that does the most work. Every major interchangeable-lens system has a copyright menu that writes your name and notice into the standard Artist and Copyright EXIF fields of every photo you take from that point on — RAW and JPEG alike, with no software, no import step, and nothing to remember.

Two things worth knowing about this setting. It applies going forward only — it will not touch photos already on the card. And it persists across owners: a secondhand body will happily keep stamping the previous photographer's name onto your work until you clear it. If you have just bought a used camera, take a frame and check the Artist field before you shoot anything that matters.

For the photos you already have — the ones shot before you found the camera menu, or shot on a phone — you need a tool that writes metadata in bulk. There are two workflows worth learning, and they suit different people.

ExifTool: free, exact, and the right answer for an archive

ExifTool is a free, open-source command-line program by Phil Harvey, and it is the reference implementation for reading and writing photo metadata. Nothing else gives you this much control. Start by looking at what a file already has:

# What does this file claim right now?
exiftool -Artist -Copyright photo.jpg

Then write the fields. The -overwrite_original flag tells ExifTool to edit the file in place rather than leaving a _original backup beside it — leave it off the first time if you want a safety net:

# One file
exiftool -overwrite_original \
  -EXIF:Artist="Jane Doe" \
  -EXIF:Copyright="© 2026 Jane Doe. All rights reserved." \
  photo.jpg

# A whole archive, recursively (-r walks every subfolder)
exiftool -overwrite_original -r \
  -EXIF:Artist="Jane Doe" \
  -EXIF:Copyright="© 2026 Jane Doe. All rights reserved." \
  ~/Pictures/archive/

# Only fill files that don't already have a notice — leaves client
# work and other people's photos alone
exiftool -overwrite_original -r -if 'not $Copyright' \
  -EXIF:Artist="Jane Doe" \
  -EXIF:Copyright="© 2026 Jane Doe" \
  ~/Pictures/archive/

A twenty-year archive should not all say 2026. ExifTool can pull the year out of each file's own capture date and build the notice from it, so every photo gets the year it was actually taken:

# Year taken from each file's DateTimeOriginal
exiftool -overwrite_original -r \
  "-EXIF:Copyright<© ${DateTimeOriginal#;DateFmt('%Y')} Jane Doe" \
  ~/Pictures/archive/

And the version that fills EXIF, XMP, and the legacy IPTC block in one pass, so that every program that looks for a copyright finds one. The CodedCharacterSet line is not pedantry: the legacy IPTC block has no inherent encoding, and without that declaration an accented name comes back as mojibake.

exiftool -overwrite_original -r -charset iptc=UTF8 \
  -EXIF:Artist="Jane Doe" \
  -EXIF:Copyright="© 2026 Jane Doe" \
  -XMP-dc:Creator="Jane Doe" \
  -XMP-dc:Rights="© 2026 Jane Doe" \
  -IPTC:CodedCharacterSet=UTF8 \
  -IPTC:By-line="Jane Doe" \
  -IPTC:CopyrightNotice="© 2026 Jane Doe" \
  ~/Pictures/archive/

Lightroom: a preset that stamps every import

If you already keep a catalogue, the better answer is to make it automatic. Lightroom Classic lets you save a metadata preset — your name, your notice, your usage terms, your URL — and apply it in the Import dialog, so every photo is stamped as it comes off the card and you never think about it again. Adobe Bridge does the same job with metadata templates, applied to a selection of files where they sit, without importing anything into a catalogue.

Free alternatives do exist and do the job properly: digiKam is a full cataloguing application with real metadata editing, and XnView MP will batch-edit a folder through a GUI if you would rather not learn a command line.

It gets deleted. Not maliciously — platforms re-encode every upload to serve it efficiently, and the metadata blocks are discarded wholesale in the process. Your copyright notice goes into the same bin as the GPS coordinates. The photo a stranger saves off Instagram or Facebook carries no author, no notice, no camera, and no date.

The uncomfortable half of this is that the platform itself received the original with everything in it — the stripping happens on the copy served to the public, not on the file sitting on their servers. Our page on which social media sites strip EXIF data has the side-by-side, and what EXIF data Facebook stores covers what a platform keeps for itself after the public copy has been scrubbed.

So where does an embedded notice actually earn its keep? On every file that is passed around in its original form — which is most of the files that matter commercially. A JPEG emailed to a client, delivered to a picture desk, handed to a printer, or submitted to a stock library arrives with your name in it. Google Images can surface creator and credit information from these fields alongside an image, and stock and editorial desks read them as a matter of routine. A file that arrives with the rights fields filled in is treated differently from one that arrives blank.

In most countries you own the copyright in a photo from the moment you take it, whether or not anything is written into the file. Embedding a notice does not create that right and it does not register it. What it does is make the claim travel with the image.

• It removes the honest-mistake defence. Someone who uses a photo that plainly says who made it cannot claim they had no way of finding out.
• It makes you contactable. A large share of unlicensed use is people who would have paid if they had known who to ask. The notice, and especially a licensing URL in the XMP fields, turns that into a lead.
• It is legally recognised. Embedded rights information is treated as copyright management information under US law, and knowingly stripping it to conceal infringement is itself prohibited — separately from the infringement. This is not legal advice, and the details vary by jurisdiction, but it is why the fields are worth filling in.
• It does not stop copying. Anyone can remove your notice with the same one-line command you used to write it, and anyone who saves your photo from a social platform gets a file with no notice in it without even trying. Treat it as a label, not a lock.

The realistic protection stack is three layers, and metadata is only the first: embedded rights fields on every file you send out; a visible credit on anything going to social media; and your original RAW files, backed up, as the evidence of authorship that no amount of stripping can touch. Metadata is the cheap layer — it costs one command and it works on every file that leaves your machine intact — but it is not the whole answer, and anyone telling you otherwise is selling something.