RSL Automates Content Licensing for Fair Compensation of Publishers+Creators (rslstandard.org)

A coalition of major publishers and tech companies launched the Really Simple Licensing (RSL) standard and a nonprofit RSL Collective to automate machine-readable content licensing and collective rights management for the “AI-first” web. RSL—an open, decentralized protocol built on RSS and Schema.org metadata—lets any site embed licensing and royalty terms (via an extension to robots.txt and RSS catalogs) that specify free, attribution, subscription, pay-per-crawl, or pay-per-inference models. It also supports encrypted licensing for paywalled or proprietary assets and can be used standalone or through the RSL Collective, which pools rights for collective negotiation and distribution of payments. Early backers include Reddit, Yahoo, O’Reilly, Medium, Ziff Davis, Fastly and Quora; development is guided by an industry TSC including RSS co‑authors. For the AI/ML community, RSL creates a standardized, automatable interface between content owners and model builders that could materially change dataset sourcing economics and compliance workflows. Pay-per-crawl and pay-per-inference introduce direct usage-based costs for training and serving models that rely on web content, while machine-readable catalogs and encryption make licensing and provenance easier to audit and enforce at scale. If widely adopted, RSL could shift negotiation leverage toward publishers, simplify licensing at internet scale (similar to ASCAP/BMI in music), and force AI firms to either pay for, avoid, or rework use of licensed web content.