Creative Commons debuts CC signals, a framework for an open AI ecosystem

Creative Commons, the nonprofit known for pioneering licensing systems that enable creators to share their work while retaining their rights, is stepping into the AI landscape with a new initiative. On Wednesday, they introduced CC signals, a framework designed to help dataset owners specify how their data can be utilized by AI systems, especially for model training. This innovative approach aims to strike a balance between the open accessibility of the internet and the escalating need for data to drive AI advancements. According to Creative Commons, continued extraction of data threatens to diminish the internet's openness, potentially leading to more websites restricting access or imposing paywalls, rather than promoting data sharing. The launch of CC signals seeks to provide a legal and technical framework for dataset sharing. This initiative is particularly timely as organizations navigate changing policies regarding AI training on their data. Companies are increasingly revising their terms of service to clarify how user data will be used in relation to AI. For example, X initially allowed third-party access to its public data for AI training but later reversed that decision. Similarly, Reddit is using its robots.txt file to block bots from scraping its data, and Cloudflare is exploring solutions that would impose fees on AI bots accessing their information. In contrast, the CC signals project puts forth a comprehensive solution that encompasses various legal enforceability options, all imbued with ethical considerations akin to the existing CC licenses that govern billions of openly licensed works online. “CC signals are designed to sustain the commons in the age of AI,” stated Anna Tumadóttir, CEO of Creative Commons. “Just as the CC licenses helped build the open web, we believe CC signals will help shape an open AI ecosystem grounded in reciprocity.” Currently, the CC signals initiative is in its nascent stages, with preliminary designs published on the Creative Commons website and GitHub repository. The organization is actively seeking public input as it gears up for an alpha launch in November 2025, which will include a series of town hall meetings to gather feedback and address questions.