Partnership for Information and Democracy and co-chaired by Luxembourg, Spain, and Ukraine
Our Senior Associate - Research and Policy Engagement, Merrin Muhammed Ashraf, participated as a speaker in the second closed-door session of the Workstream on Safeguarding Access to Reliable Information in the Age of AI, organized by the Partnership for Information and Democracy and co-chaired by Luxembourg, Spain and Ukraine on 18 June 2026. The online working group meeting brought together approximately 50 participants, including representatives of signatory states and civil society organisations, to examine policy and legal approaches for safeguarding the viability and visibility of media in the age of artificial intelligence.
During the session, Merrin presented IT for Change's issue brief, "Governing AI for the Cultural Commons: Beyond Intellectual Property," developed in collaboration with the Subgroup on AI, Intellectual Property and Culture under the UNESCO Global CSOs and Academic Network on AI Ethics and Policy. The presentation explained how AI systems increasingly rely on open repositories, archives and collaborative knowledge infrastructures such as Wikipedia, often without attribution, reciprocity or meaningful value sharing. It argued that existing open licensing frameworks, including Creative Commons licences, were designed to facilitate reuse under defined conditions but have become increasingly difficult to enforce within contemporary AI training pipelines.
Merrin further argued that the current intellectual property regime enables AI extractivism in two principal ways: first, through legal exceptions such as fair use, text-and-data mining provisions and open licensing arrangements that permit large-scale extraction of knowledge resources; and second, by allowing the value generated from these resources to be enclosed within proprietary AI models protected through patents, trade secrets and restricted access. Moving beyond narrow copyright reform, the presentation advocated community-centered licensing models with differentiated access conditions, stronger accountability to source communities and benefit-sharing frameworks, particularly for communities whose cultural knowledge contributes to AI systems.