A new open licensing standard called Really Simple Licensing (RSL) has been introduced to help web publishers manage and monetize AI scraping of their content, with support from major publishers including Reddit, Yahoo, and Medium.

The RSL protocol involves publishers including licensing terms in their 'robots.txt' files, making it easier to identify permissible data for AI training, while a collective licensing organization, the RSL Collective, negotiates terms and collects royalties.

Led by industry veterans Eckart Walther and Doug Leeds, the RSL Collective promotes RSL as a scalable business model similar to rights organizations in the music industry and is developing gatekeeper technology with Fastly to manage bot access based on licensing.

The success of the standard relies heavily on cooperation from AI companies; without their buy-in, enforcing licensing terms could prove challenging.

The group believes RSL will be legally enforceable, citing recent legal cases like Anthropic's $1.5 billion settlement to highlight the financial risks for AI companies that train on unlicensed data.

Technical enforcement is being facilitated through partnerships with cloud services like Fastly, which can act as gatekeepers to block or permit web crawling based on licensing compliance, with future involvement from companies like Cloudflare.

The standard aims to create a sustainable and fair ecosystem that balances AI innovation with content creator rights, potentially influencing future regulations and industry practices.

Adoption by AI companies is crucial but challenging, as many developers have historically ignored licensing instructions; collective action and enforcement mechanisms are being developed to address this.

Major AI firms have often disregarded site restrictions, but RSL seeks to unify licensing frameworks to change this behavior.

A significant challenge is the difficulty in tracking when training data is ingested and attributing royalties, especially when training isn't logged or transparent.

This move reflects a broader effort to establish clear legal frameworks for AI training data, addressing concerns over data use and copyright.

While the RSL system is designed to be legally enforceable, online content licensing for AI training remains legally grey, with ongoing lawsuits involving platforms like Reddit and Getty Images.

The adoption of licensing standards like RSL is vital in the evolving AI landscape to ensure transparency, fair use, and proper compensation for content creators.