Really Simple Licensing (RSL)

Really Simple Licensing (RSL) is an open, XML-based standard that lets publishers state in machine-readable form how AI systems may use their content (for training, as input to answers, for indexing or for search) and what compensation or attribution each use requires. It extends the idea of robots.txt from 'may you crawl this' to 'under what license may you use this'.

RSL was launched on September 10, 2025 by the RSL Collective, a nonprofit rights organization co-founded by Eckart Walther, a co-creator of RSS, and Doug Leeds, former CEO of IAC Publishing. Reddit, Yahoo, Medium and Ziff Davis were among the launch supporters. On December 10, 2025, the RSL Technical Steering Committee published the RSL 1.0 specification with the status of Recommendation, meaning stable and ready for deployment. Its editors include people from Condé Nast, Ziff Davis, Yahoo, Automattic, O'Reilly Media, Fastly and Schema.org.

An RSL license is an XML document that pairs permitted or prohibited uses with payment terms. The core usage vocabulary, which incorporates Cloudflare's Content Signals terms, includes:

• ai-train: training or fine-tuning models.
• ai-input: using the content as input to generate answers, including retrieval-augmented generation, grounding and AI search summaries.
• ai-index: storing it in an AI system's index or retrieval database.
• search: building a classic search index with links and short excerpts, explicitly excluding AI-generated summaries.
• ai-all: any AI use.

Payment types range from free and attribution (visible credit and a working link) to crawl (a fee per crawl), training, use (a fee each time the content contributes to an AI output), subscription and purchase. Prohibitions win over permissions when the two conflict.

Crawlers find the license through any of several equivalent mechanisms: a License: line in robots.txt, an HTTP Link header with rel="license", an HTML link element, an RSS module, or metadata embedded in files such as EPUBs. Three optional protocols handle enforcement for those who want it: the Open License Protocol (an OAuth 2.0 extension for obtaining licenses), the Crawler Authorization Protocol, and an encrypted media standard. Platforms can also point RSL at their own enforcement, and the specification names CDNs and pay per crawl services as valid ways to collect fees. The RSL Collective runs a collective licensing platform on top of the standard, offering 'pay-per-output' royalties similar to how music rights organizations work.

Like robots.txt, RSL is a declaration. It only binds crawlers that read and honor it, so enforcement in practice depends on CDN rules, Web Bot Auth identification and contracts.

The usage split is what makes RSL relevant to GEO. A brand that wants to be cited can permit ai-input and search while prohibiting ai-train or attaching a fee to it. Prohibiting ai-all on marketing pages would ask AI search systems not to use your content in answers, which is rarely what a company selling a product wants. Before publishing a license, map which page types exist to win AI search visibility and which are products in themselves, write terms for each group, and then compare crawler behavior in your logs before and after. RSL is one of several overlapping signals alongside TDM rights reservation and content signals, and the same intent should be expressed consistently across all of them.