What is Really Simple Licensing (RSL)?

Really Simple Licensing is an open standard for publishing machine-readable terms for your content, including how it may be used and what that use costs. A small file on your own domain identifies the content, the uses you permit or prohibit, and how you expect to be paid. robots.txt tells a crawler whether it may fetch a path. RSL adds the licensing terms and price, so an operator can see the conditions for using what it fetches.

An open standard, not a product

RSL 1.0 is a published specification maintained at rslstandard.org. Anyone can implement it. No company owns it, there is no licence fee for the standard and no platform to join. You publish a file on your own domain.

Previous attempts asked publishers to put their content with an intermediary first. RSL asks for 15 lines of XML on the site you already run. Supertab is one of several implementers. You can use the standard without any of us.

What an RSL file can actually say

RSL uses defined elements and tokens to describe terms. <content> identifies what is covered, down to a URL path. <permits> and <prohibits> specify allowed and prohibited uses: ai-train for training, ai-input for a model reading your page to answer a live question, and ai-index and search for indexing. User tokens also let you set different prices for commercial and non-commercial users.

The <payment> element sets the cost. Its types include training, crawl, use, subscription, purchase, attribution and free. One file can set different prices for training and live retrieval of the same article.

Who publishes it and who reads it

You publish the file. AI companies, crawler operators, agents and the tools built around them read it. A predictable location on your domain makes it discoverable without a directory, marketplace or introduction.

RSS used the same approach: a file on your domain that other systems could read without signing an agreement. RSL applies that approach to publishing permissions.

Why terms had to become machine-readable

Software uses your content at a scale that people cannot review request by request. A crawler cannot evaluate an English terms-of-service page as a licence. At 10 million requests a day, referring each decision to a person is not an option.

The 2019 European copyright directive lets rightsholders refuse text and data mining on the condition that the refusal is expressed in a form machines can process. Courts have begun testing what that means. As far as we can see, a machine-readable file fits the form the law was trying to describe.

What RSL doesn't do

RSL does not block requests. A crawler that ignores your terms can still receive the page, just as one that ignores robots.txt can today.

Enforcement requires a separate layer, including tokens and verification. Keeping it separate makes the standard easier to adopt. RSL spread quickly because it requires very little from publishers. Trying to make one standard handle both the licence and access enforcement would have prevented that adoption.

‍

Written by the Supertab Team