# What is the Protege AI data marketplace?

> The Protege AI data marketplace sells licensed training data in health, video and audio, and pays partners a share. What is public, and what is not.

By Matt Burns, Vaultfish. Canonical: https://vaultfish.com/answers/protege-ai-data-marketplace  
Published: 2026-08-03  
Updated: 2026-10-14

## The short answer

The Protege AI data marketplace is a site where AI builders license training data. Partners get a share of the money each time it is used. It lists health, video, audio, speech and spatial data. Its pages never mention workplace chat or email. The share is not public.

## What does the Protege AI data marketplace do?

[Protege](/buyers/protege) (withprotege.ai) calls itself "a curated marketplace and compliance layer." Model builders "discover, request, filter and combine datasets" there ([SiliconANGLE](https://siliconangle.com/2026/01/08/protege-raises-30m-grow-governed-marketplace-ai-training-data/)).

Its homepage tells owners to "monetize your existing data assets or content" while keeping "clear rights protections and provenance."

The domains it names are healthcare, video, audio and speech, and spatial and physical data. It also lists pre-training, fine-tuning and evaluation sets. A SOC badge shows in its footer.

> **Protege lets AI developers license data, and it pays the owner each time the data is used.** It's like a library that pays the author every time a book is borrowed.

## How does Protege pay its data partners?

Its funding coverage says partners receive "revenue share payouts ... with each use." CEO Bobby Samuels told EO Magazine it uses a royalty-based structure: "You make money when we make money" ([EO Magazine](https://www.eomag.io/article/protege-bobby-samuels)).

The share itself is not public. Its video provider page lists two case studies of what a broadcaster and a sports distributor received. Protege reports both itself. I leave the amounts out, because one company's result predicts nothing for yours.

## Who can license data to Protege?

The video page says partners typically own content outright or hold "global all media or similarly broad rights." It states no minimum. Its process has four steps, from discovering a library to receiving payments.

Figure: steps Discover and structure -> Prepare -> Monetize. Protege's process runs from a library review to revenue share payments.

## How much has Protege raised?

SiliconANGLE reported a $30M Protege round led by Andreessen Horowitz in Jan 2026. It put the total at $55M, including a $25M Series A from August 2025. EO Magazine puts the total at $65M.

The two press figures differ, so quote the one you can name.

## Is Protege an option for Slack, email or CRM records?

Protege's pages that we read do not name workplace chat or company records. Ask Protege directly before you assume it covers yours.

_A library that pays the author on every loan. The rest of us just get a late fee._

Our [buyer directory](/buyers) and [deals tracker](/deals) cover the players that do list company data. The [rights check](/rights-check) covers what you need to hold before you speak to any of them.

Our only referral relationship is with micro1, an AI data company. It may pay us a fee if a company we introduce goes ahead with it. You pay nothing extra. [How we are paid](https://vaultfish.com/disclosure)

Have a lawyer read any contract before you sign. They like that.

_Vaultfish helps you run the process. We are not lawyers and this is not legal advice. If you need counsel who knows this area, we can point you to one._

## Sources

- [SiliconANGLE](https://siliconangle.com/2026/01/08/protege-raises-30m-grow-governed-marketplace-ai-training-data/)
- [EO Magazine](https://www.eomag.io/article/protege-bobby-samuels)
- [Protege homepage](https://withprotege.ai/)
- [Protege video data providers](https://withprotege.ai/data-provider/video)