RSL & the future of copyrighted “creative content”
Finally an answer to AI-crawlers & scrapers
Not too long ago, I had a discussion about the theft of millions of creative thoughts, writings, art, materials etc from AI-companies. “Wouldn’t it be possible…”, I said, “…to install a small piece of code into any place of a website (e.g.: the “robots.txt” file) to stop or block AI-web-crawlers from harvesting the “cultural content” from the page?”
What at that time seemed to be a cool project for one of my tech-acquaintances, became now reality much faster than anticipated. Well, at least in a way — but different. Better actually. RSL is new, but its goal is not to stop the crawlers, but rather to clearly let them know, that they are trespassing on someone else’s property — and thus have to ask permission to pluck some flowers from the garden.
NOTICE: all crawlers and bots are strictly prohibited from using this # content for AI training without complying with the terms of the RSL # Collective AI royalty license. Any use of this content for AI training # without a license is a violation of our intellectual property rights.
“Don´t sneak in here and steal my art!” — this is a clear message. Legally, this is a strong signal, that this content is under copyright — and therefore an agreement of any kind shall be necessary. Cloudflare acted on this already in July 2025 (“stop the scraping”) and argued for a “permission-based approach”.
This is not a flimsy matter. This is about money.
“Royalties” are an essential part of every creators income mix — and copyright has been a highly contested subject since the dawn of “technical reproduction possibilities”. In the digital age, new forms of copyright protections and collective rights organizations had to be established — just to give a little compensation to the original creator of whatever is used, utilized and too often abused.
Of course, many of these tiny pieces of compensation are much too small. Spotify is paying fractions of cents / pennies to the average musician — but even this has been a long fight originating in the music download controversies in the early 2000´s. And now, 25 years after Napster “… nearly killed the music industry.”, AI could eradicate a lot of the current creator “industry” — when this in reality undefinable business field will not quickly adapt to the massive changes.
”…June 1999 … Napster, the brainchild of Shawn Fanning, linked computers and allowed users to access each other’s mp3 audio files. Fanning met an ambitious teen, Sean Parker, on the webchat channel w00w00, and together they triggered a series of events that brought the record business to its knees by making music discovery instant — but payment optional…”
Today, we are in a similar disruptor situation with the LLMs, which have been (and partly still are) “trained” by billions of webpages without consent of the owners. More and more complaints and law suits are happening, so that the appearance of a non-profit digital rights organization could not be more timely.
RSL stands for “Really Simple Licensing” and promises to be:
a) a standard; just a simple piece of code, alerting any crawler, that they need to ask for permission to use the (creative) content of the site — and
b) a basis for a non-profit organization (the “RSL collective”) to collect royalties from AI-usage.
”…According to RSL co-founder Eckart Walther, who also co-created the RSS standard, the goal was to create a training-data licensing system that could scale across the internet. “We need to have machine-readable licensing agreements for the internet … That’s really what RSL solves…” (TechCrunch)
Eckart Walther explained the concept in detail during the recent MEDIUM-Day in a 30-min session (which will hopefully become available on Medium’s YouTube channel). This was very insightful and the concept is quite a no-brainer for most creators, the vast majority of us much too weak as individuals to enforce any digital rights.
I remember 15 years ago, when the German music rights association GEMA had a long conflict (2009–2016) with YouTube and many videos were geoblocked in Germany, because YouTube did not offer enough compensation for the musicians which appeared in any home-made soundtrack. The situation can loosely be compared to today’s music usage in TikTok video creations… how much money does any musician actually really get from the plays on that platform?
3 years after the sudden & shocking appearance of ChatGPT in November 2022 (and after much handwringing and complaints in the creative & culture field), there are finally signs of possible solutions, which may force some reactions from the cold tech-AI-LLM-behemoths — at least. Yet, there seems to be also options of possible cooperations, creating maybe some understanding.
”The role of technology is to be our assistant, not our replacement.”
( during Medium Day 2025)
RSL was just introduced 2 weeks ago (the wikipedia page (edited by Funcrunch) was created 2025–09–10), but it gains already some momentum, as it can be used indeed really simple.
Of course, tech companies and platforms may just evade the responsibility by opening the floodgates to ever more AI-slop. (interestingly, that wikipedia page was just created a year ago on 2024–09–28 by Wikipedia editor Jenny8lee)
However, original thought & authenticity can never be replaced … hopefully!
(I may explore this thought a bite (!) more soon)
For now, I have decided to use RSL on my own websites.
“..So we’re happy, finally, to support a formal, standardized way to for all content owners to tell AI companies the rights and restrictions on your content…”
👉🏽 read more details about how MEDIUM has tested solutions in the last years in the RSL introductory post
How RSL Works
Key capabilities of RSL include:
- A common, extensible vocabulary that lets online publishers define licensing and compensation terms, including free, attribution, pay-per-crawl, and pay-per-inference compensation.
- An open protocol to automate content licensing and create internet-scale licensing ecosystems between content owners and AI companies.
- Creating standardized, public catalogs of licensable content and datasets through RSS and Schema.org metadata.
- An open protocol for encrypting digital assets to securely license nonpublic, proprietary content, including paywalled articles, books, videos, and training datasets.
- Support for collective licensing through the nonprofit RSL Collective rights organization or any other RSL-compatible licensing server.
via RSL orginal press release: https://rslstandard.org/press/rsl-standard
#tech, #licensing, #copyright, #AI, #web
