Really Simple Licensing (RSL) for AI
 Author
Updated
See also
IntroductionReally Simple Licensing (RSL) https://rslstandard.org is an open standard that allows publishers to set machine-readable licensing terms on their web pages for AI crawlers. AI is changing the way information is found, exchanged and repurposed on the web; however, most large language models are built, at least in part, on content scraped from websites without permissions or compensation for creators. RSL aims to change this and to bring more order, fairness, and sustainability back to the Web re: intellectual property. RSL is a machine-readable format to set explicit terms for how a creator's or publisher's content can be used by AI systems. Instead of agreeing/disagreeing to specific AI bots, websites can add specific licensing and royalty terms to their robots.txt file; and, embed terms in online books, videos, and training datasets as well. RSL makes content licensing simple and enforceable. RSL may simplify how publishers license their digital content. Instead of negotiating separate agreements with each entity seeking to use a publisher’s content (or having no agreement at all), publishers can use RSL to clearly state their licensing preferences in a machine-readable format. Time will tell whether this RSL standard will take hold by content creators, and integrated into the broader legal and social contexts. History and Origins of RSLRSL was launched on September 10, 2025, recognizing that AI scraping was creating legal and ethical problems for creators, intellectual property owners and publishers. The RSL initiative is led by Eckart Walther, co-creator of RSS, and Doug Leeds, former CEO of Ask.com. Their plan is to build a licensing layer that mirrors RSS’s simplicity but responds to today’s AI-driven challenges. The non-profit RSL Collective will manage the standard, and bring together publishers and creators to protect their rights and negotiate fair compensation. This coalition underscores the urgency of publishers to regain control and monetize their intellectual property. Really Simple Licensing (RSL) protocolReally Simple Licensing (RSL) https://rslstandard.org is an open, XML-based document format for defining machine-readable licensing terms for digital assets, including websites, web pages, books, videos, images, and proprietary datasets. Really Simple Licensing (RSL) builds on the robots.txt protocol and publishers can embed licensing terms directly into files. Instructions could include charges per crawl, subscription fees, and payment every time an AI model references content. Technical implementation will require publishers to add XML-based licensing terms to their robots.txt files. AI crawlers read these terms and, theoretically, comply with them. The RSL Collective is modeled after music licensing groups like ASCAP and SOCAN (Canada), and handles the negotiations and royalty distributions. RSL requires the following:
Librarian perspectivesLegal, technical, and ethical enforceability: early discussions with librarians about AI licensing focus on whether terms are enforceable in practice (robots.txt is one example of a weak technical barrier historically respected by some, ignored by others). In fact, no major AI firms (e.g., OpenAI, Google) have committed to RSL yet. In the absence of a legal mandate, and copyright reforms, RSL may be slow to catch on, similar to robots.txt. Librarians, who manage large digital library collections, may see RSL as a non-starter if crawlers get around it. One librarian said, "Either training is not considered copying in law... or we require a major legal overhaul re: AI...we don't need quick fixes". Librarians have long promoted open ecosystems such as open-access journals, open textbooks and fair dealing/use doctrines; and open principles of Creative Commons licensing. Will RSL skew towards restricting access, especially public access? RSL may exacerbate closed silos, where AI tools (used by researchers and universities) will prioritize paid content from multinational publishers. Some libraries' digitized archives will require "pay-per-inference fees" (fees to compensate publishers when AI crawlers or system actually use their content in generated answers). RSL may inadvertently minimize public-domain or nonprofit content, creating a range of biases due to dominant mainstream perspectives. Benefits for Publishers and Content Creators
Benefits for End Users (Consumers of Content and AI)
References
Disclaimer
|