John Boyden of the RSL Collective has proposed a machine-readable, open-web standard designed to allow news publishers to define the terms of employ and compensation for their content when utilized by artificial intelligence systems.
Speaking at a recent WAN-IFRA webinar titled “The Power of the Collective: Navigating AI Partnerships & Content Licensing,” Boyden stated that AI companies are currently consuming publisher content on a massive scale without providing compensation. He argued that this practice is accelerating the loss of digital revenue and eroding the relationship between publishers and their audiences by undermining traffic-based revenue models.
According to Boyden, while a small number of high-profile licensing deals have been reached, the vast majority of publishers currently receive effectively zero compensation for the use of their intellectual property in AI training and output.
A Framework for Collective Licensing
The RSL Collective, a non-profit organization, aims to reverse this trend by providing a mechanism for publishers to reclaim lost digital revenue through collective licensing. The organization’s strategy rests on three primary components: a rights framework, a collective licensing framework, and a per-use royalty model specifically tailored for AI.
Boyden identified a significant friction point for AI developers: the fragmentation of rights. Because content is owned by millions of individual publishers without a unified system for granting permissions or determining a consistent price, AI companies lack a scalable way to pay for ongoing use. Boyden described this as a $100 billion opportunity for publishers, noting that while spending on AI technology is exploding, the economy for open-web publishers has remained stagnant.
The RSL Collective proposes to capture a percentage of the revenue AI companies generate from their products and earmark those funds for the content creators. The organization already represents a diverse group of publishers and creators, including Yahoo, Reddit, Wikipedia, USA Today, Torstar, Ziff Davis, and People Inc.
Technical Integration and Revenue Model
To ensure the system is scalable, RSL is collaborating with infrastructure providers Cloudflare, Akamai, and Fastly. These companies have agreed to use the RSL standard to validate AI crawlers, providing a technical layer to the rights assertion process.
The organization operates as a non-profit, employing a revenue model similar to the music industry’s ASCAP. RSL retains a percentage of the licensing revenue that flows through its system to cover operational costs, including technology, salaries, and office space, without generating a profit.
Boyden emphasized that the collective is non-exclusive and requires no membership dues or long-term commitments. Publishers are permitted to join or depart the collective monthly and remain free to negotiate independent, direct deals with AI companies if they believe they can secure more favorable terms.
The initiative is positioned as a benefit to AI companies as well, offering them a single point of negotiation rather than the require to execute tens of thousands of individual contracts.
The discussion on AI licensing is scheduled to continue at several upcoming industry events, including the Frankfurt AI Forum on April 13-14, the World News Media Congress from June 1-3, and AI Forums in Kuala Lumpur and Hanoi in August.
