In June we kicked off a public feedback period on our proposal for CC signals. CC signals is a preference signals framework designed to sustain the commons and ensure the continued sharing of knowledge in the age of AI.
The goal is to give holders of large datasets a way to set criteria for how their data may be used within AI training models. To give an example, a dataset holder may wish to require that any AI training that uses their data gives credit back to the original source (e.g. attribution), or that the resulting AI model is open. Like the CC licenses, CC signals builds on the idea of ‘some rights reserved’ and that creators and knowledge holders deserve meaningful choices in how their work is used. You can learn more on our website.
Since our kickoff event, we have been listening closely to feedback. We heard from hundreds of creators, librarians, technologists, legal experts, and open advocates. We asked for feedback and you delivered! Your voices – supportive, skeptical, frustrated, or curious – are essential in shaping how CC signals develops. We’d like to summarize what we heard and how this feedback is being incorporated and addressed.
What We Heard
Across the conversations, several themes emerged:
Concerns that CC is prioritizing AI companies over creators. A recurring concern is that CC signals seem to give legitimacy to AI training without doing enough to protect creators.
Confusion and disagreement about the CC licenses and AI training. We heard frustration that the CC licenses are not being interpreted or enforced in ways that some creators expected.
Strong calls for opt-outs. Many wondered why the draft CC signals did not include an opt-out option.
Asking politely for AI developers to give back in exchange for datasets is not enough. We heard doubts that CC signals would work in practice, given the widespread evidence of AI companies ignoring copyright, licenses, and even technical protocols like robots.txt.
Broader critique of AI’s role in society. There is a spectrum of views on AI across the CC community. Many of you stand firmly at the anti-AI end. For these voices, no technical framework, like CC signals, feels adequate without stronger laws and regulations.
We haven’t been clear on who this tool is meant to serve and the use cases it is meant to address. Naturally, the needs of an individual creator, like an artist, are quite different from those operating at an institutional or collective level. We heard loud and clear that CC signals, as currently conceived, does not meet the diverse needs of individual creators.
Requests for clarity. Many asked for more details about implementation and interoperability, including our long-term vision for CC signals as part of our broader mission.
We understand how deeply personal these issues are for many of you, especially artists and creators who feel their work is being taken without consent and are looking for ways to fight back. That frustration is real, and we take it seriously.
What We’re Doing Next
✔️Improving clarity around CC’s position. We know many of you are worried that CC has “taken sides” or is being influenced by AI companies. We want to be clear: the driving motivation of CC signals is to defend and sustain the commons by developing practical tools for knowledge holders. Going forward, we will aim to clarify our guiding principles and positions in ways that translate to product decisions.
✔️Strengthening messaging and education. We are committed to expanding resources on how the CC licenses and CC signals could interact, examples of how signals could work in practice, and deeper dives into questions of copyright within the AI landscape. If you haven’t already, take a look at our legal primer on understanding the CC licenses and AI training. The better informed the CC community is about AI and the commons at large, the more effective we can be as a community to defend the commons.
✔️Clarifying the use cases for CC signals. This phase of CC signals is designed to serve large and open dataset holders, not the individual creator. Your feedback helped us recognize that this focus was not easy to square with our decision to leverage technical protocols used by anyone with a website. As a result, the target audience for CC signals was not clear. As we decide on next steps in product development, we plan to focus on specific use cases to put our goals and objectives into practice.
✔️Deepening global engagement and inviting stakeholders into product development. We plan to continue conversations with diverse audiences to inform the future of CC signals through an iterative process. The rest of this year will be focused on exploring and testing possible integrations of CC signals with pilot adopters. From this, we hope to extrapolate findings as we explore wider adoption of CC signals in the future.
✔️ Maintaining transparency in development. Our GitHub repository will stay open and up to date. We are creating a roadmap that will be shared publicly and will provide consistent updates (either on the blog or via a virtual town hall) on our progress. This feedback loop is not over; it will be built into how CC signals will evolve.
Looking Ahead
The future of the commons depends on tools that reflect shared values of openness, fairness, and agency. We know many of you remain skeptical.
CC signals is not final. It is an experiment in building a new layer of choice in an age where the rules are rapidly shifting. We will keep listening, adjusting, and collaborating until we arrive at something that genuinely serves the commons.
Thank you to everyone who took the time to write, question, challenge, and support us. Please stay engaged. Together, we can ensure that Creative Commons continues to stand where it always has: with the community, for the commons.
Posted 27 August 2025