Dealroom.co: content licensing and data marketplace startups like Pip Labs and ProRata have raised $215M since 2022 to help artists sell content for AI training
New crop of companies are offering artists ability to sell content to big tech firms to train AI See also Mediagazer
Context & Ripple Effects
This funding tally puts Pip Labs and ProRata in an emerging intermediary layer between creators and AI developers. The market has been taking shape alongside reported training-data deals that assigned explicit prices to images, video, and long-form footage, turning content rights into a more legible commercial input for model development.
ProRata had already raised a Series A around a model to share chatbot revenue with content owners and later signed media licensing agreements, suggesting that marketplaces are being paired with attribution and downstream-revenue mechanisms rather than one-off data sales alone.
First-order effects
- Artists and other rights holders gain more potential routes to package and sell training rights, while startups such as Pip Labs and ProRata have capital to operate the matching, licensing, and payment layer.
- AI companies get additional specialist counterparties for sourcing licensed material, rather than negotiating every arrangement directly with individual creators or publishers.
Second-order effects
- Marketplace operators will need to differentiate on rights verification, pricing, and reporting: ProRata's earlier revenue-sharing approach for content owners illustrates that licensing terms may extend beyond an upfront dataset transaction.
- Direct licensing by publishers and creators becomes more consequential as marketplaces compete to aggregate supply; buyers may compare the cost and provenance of intermediary-sourced content against bilateral deals.
Third-order effects
- If these intermediaries gain adoption, AI-training content could evolve from irregular rights negotiations into a more standardized market with reusable licensing and attribution infrastructure.
- The key unresolved question is whether revenue-sharing and provenance systems can become trusted enough to distribute AI-derived value across many rights holders, rather than concentrating bargaining power with the largest content owners and model developers.
The trend: AI development is creating a commercial rights-and-data supply chain in which specialized platforms seek to make licensed content purchasable, traceable, and compensable at scale.