Can a consumer brand or photo studio sell its images and files to AI companies?

Sometimes. A brand or studio that has shot thousands of products holds a large, well-labeled image library. Whether it can be licensed depends on paperwork rather than on the pictures: who owns each photograph, whether people in them signed releases, and what the original license allows.

Contact usWhat's my data worth?

What a brand or photo studio has that is licensable

Product photography with catalog dataImages of items on plain backgrounds, tied to a product number, color, size and description. The pairing is what makes them useful.
Shot lists and creative briefsThe plan for a shoot, the angles asked for and the notes on what the finished set should show.
Retouch notes and edit historyThe instructions given to a retoucher, the before and after versions and the reason for each change.
Campaign files and style guidesLayouts, copy decks and brand rules that explain how an image is chosen and cropped for a given channel.
Alt text, captions and metadataDescriptions written for accessibility, search or a catalog, which are labeled text for each image.

Where the records live

Brands keep final images in a digital asset management system or a cloud drive, and catalog fields in a product information tool. A studio often holds raw files and layered edits on its own servers and sends the finished set to the client.

We ask who holds the originals, how many years of images exist and whether the metadata travels with each file. A large library with no labels is worth much less than a small one with good descriptions.



People in the picture

Images with a recognisable person depend on a signed model release, and releases are often limited by use, time and place. In some jurisdictions a person's likeness and biometric data have extra rules. Product-only images with no people avoid this question, which is why they are usually considered first.


What stays out for a brand or photo studio

Images with people and no usable releaseIf a signed release covering this use cannot be found, the image is left out.
Photos owned by someone elseFreelancer work, stock licenses and agency photography stay with their owners unless they agree.
Third-party trademarks and artworkProps, packaging or backgrounds that show another company's mark or an artist's work are checked or removed.
Unreleased products and campaignsImages of items not yet on sale are held back until after launch.

More on this: how personal details are removed, who carries the risk and how you get paid.


A sample, step by step

A home goods brand with sixty staff has eight years of product photography in a digital asset system, each image linked to a catalog record. The brand's counsel checks which photographers assigned their rights. The first package is product-only images from staff and assigned shoots, with catalog text and shot lists, and no images showing people. The brand reads the sample and decides whether to proceed.


Questions owners ask about this

Can we license photos a freelancer shot for us?

Only if the agreement gives the brand the right to do so or the photographer agrees in writing. A license limited to marketing use does not cover it.

Do old model releases count?

They may, but many are narrow and predate AI uses. Each release is read for its stated uses, and unclear ones mean the image is omitted.

What about images we edited with AI tools?

Output from some tools has limited or uncertain copyright, and tool terms differ. These images are separated and checked on their own.

Is a studio's raw file archive useful?

It can be, because raw files with edit history show the full process. They are also the biggest in size and need clear client permission first.

Do labels have to be in English?

No, but consistent labels matter more than the language. A mixed library is usually grouped by language when assessed.

Didn't find your question? Ask us directly.


Keep reading

Records to look at first

  • Project and task history: Jira, Asana, Linear, Monday, Trello, ClickUp and Airtable comments and status changes, not task titles.
  • Client deliverables and drafts: Each draft beside its final version with the comments that changed it, kept only where the client agrees.
  • Google Drive export: How to keep the comments and the version history when documents leave Google Drive.
  • Labeled tables and datasets: Rows, a result column and a data dictionary: when a table can be licensed.
  • Shopify data export: How to get orders, products and customers out of Shopify, why order rows are not orders, and what shopper details come out first.
  • Image collections: Product shots, condition photos, scans and design archives, sorted by who owns each picture and what note came with it.

Other industries