Creative Commons shares how the rise of generative artificial intelligence has affected cultural heritage institutions and open sharing.
Cultural heritage institutions (CHIs) are tasked with the public mission of providing open access to their collections. This mission is driven by the understanding that heritage defines shared humanity and that open access yields significant societal benefits. As stewards of cultural knowledge, CHIs play a vital role in expanding and safeguarding the cultural commons, enabling uses that benefit all humanity.
The emergence of generative artificial intelligence (AI) has significantly impacted CHIs, as their curated content is increasingly utilized for AI development. This raises crucial questions about the conditions, norms, and expectations governing the use of openly licensed and public domain heritage materials as AI training data. CHIs face challenges including increased bot traffic straining infrastructure, a lack of reciprocal value sharing from profit-driven AI companies, and commercial AI practices that often conflict with commons values (e.g., lack of transparency, attribution, decontextualization, and enclosure of content), which collectively erode trust amid high legal uncertainty around copyright and CC licenses.
In response to AI scraping, CHIs are faced with a dilemma: allow unrestricted extraction or implement broad restrictions, leading some, like the National Library of the Netherlands and Trove, to limit access. This defensive enclosure creates a paradox where institutions dedicated to sharing end up locking away heritage, undermining years of open heritage efforts and the principles of the Open Heritage Statement. A more sustainable equilibrium is needed beyond this reactive binary, as complete closure is inefficient and detrimental to public interest.
Recognizing CHIs' desire to share responsibly, Creative Commons is developing 'CC Signals,' a new tooling designed to restore agency to data stewards while maintaining the open nature of heritage. CC Signals aims to strike a balance between preserving access to valuable heritage and ensuring CHIs remain active participants in shaping the AI ecosystem. By establishing standardized, prosocial conditions for AI access and reuse, potentially leveraging contract and copyright law, CC Signals seeks to provide a middle path between 'free for all' and 'defensive enclosure,' ensuring public interest reuse and fostering a thriving commons.
To collectively shape an AI ecosystem that benefits open heritage, individuals are encouraged to engage. This includes signing the Open Heritage Statement to support ethical AI, learning more about and getting involved with the CC Signals project, contributing to the development of open infrastructures and prosocial technologies aligned with open values, and investing in open infrastructure to sustain these efforts.