The Emergence of a Data Winter: The Growing Enclosure of Data at a Time of Rapid AI Advances

Paper by Stefaan Verhulst: “We are entering a data winter-a period marked by the growing inaccessibility of data critical for science, governance, and innovation. Just as previous AI winters saw stagnation in artificial intelligence research, today’s data winter is characterized by the enclosure of data behind proprietary, regulatory, and technological barriers. This contraction threatens the capacity for evidence-based policymaking, scientific discovery, and equitable AI development at a moment when data has become society’s most strategic resource. Eight interrelated forces are driving this decline: government open data cutbacks, shifting institutional priorities toward risk aversion, generative AI-induced data hoarding, research data lockdowns, scarcity of high-quality training data, risks of synthetic data substitution, geopolitical fragmentation, and private-sector closure. Together, these trends risk immobilizing data that could otherwise serve the public good. To counteract this trajectory, the paper proposes five strategic interventions: (1) shifting norms and incentives to treat data as essential infrastructure; (2) translating openness commitments into enforceable action; (3) investing in professional data stewardship across sectors; (4) advancing governance innovations such as digital self-determination and social license mechanisms to restore trust; and (5) developing sustainable data commons as shared infrastructures for equitable reuse. The technical means exist, but the collective will is uncertain. The paper concludes by arguing that the coming years will determine whether society builds an open, collaborative data ecosystem-or succumbs to a fragmented, privatized data order…(More)”.