Reddit, Yahoo, Medium and more are adopting a new licensing standard to get compensated for AI scraping (www.engadget.com)

A coalition of major web publishers—including Reddit, Yahoo, Medium, and others—has introduced the Really Simple Licensing (RSL) standard, a new framework designed to establish clear licensing terms for AI scrapers using their content. By extending the familiar robots.txt protocol, RSL allows publishers to specify usage models such as free access, attribution, subscription, pay-per-crawl, or pay-per-inference—where AI companies pay only when content actually informs a model’s output. This innovation aims to address mounting publisher concerns about unauthorized AI training on their data by providing a scalable, standardized approach to compensation and content usage rights. The RSL initiative is further supported by the RSL Collective, a nonprofit acting as a rights manager—similar to those in the music industry—that seeks to set fair market prices and bolster publishers’ negotiating leverage in the AI era. Technical enforcement remains challenging since RSL cannot directly block bots; however, partnerships with infrastructure providers like Fastly offer potential gatekeeping mechanisms to regulate access. While it remains uncertain if AI companies will universally honor the new standard, proponents highlight incentives for legal compliance, such as simplifying licensing at scale and improving the quality of AI-generated answers by relying on legitimately licensed, high-quality sources. The introduction of RSL represents a significant step toward bridging the content licensing gap in AI training, with implications for fair compensation, legal clarity, and sustainable AI development.