FACT SHEET: CommonsCasting 2035: Exploring a Future Intellectual Property, Data, and Knowledge Commons for Public AI Infrastructure

The proposed project aims to investigate the potential for a future intellectual property (IP), data, and knowledge commons to support public infrastructure for research, information management, and innovation. Artificial intelligence (AI) capability is expanding at a rate incommensurate with the rate of adaptation of our IP and data infrastructure. Interest in productivity gains is now being replaced by actionable concern, manifesting in a variety of initiatives to address risks such as IP misuse, data-poisoning threats, and over-consolidation of AI capability. Resource-commons models have been proposed as a foundation for shared infrastructure, but implementation requires careful consideration of complex interconnections of incentives, rights, duties-of-care, and BOLTS (Business, Operations, Legal, Technical, Social) requirements. Proposed is the planning of a yearlong program to convene an interdisciplinary, multi-organization initiative to catalog resources, identify risks, opportunities, requirements, and remedies. As part of this initiative, participants would engage in a "ThreatCasting" exercise to explore and clarify what a functional future IP, data, and knowledge commons (and its absence) might look like. ThreatCasting methods will be adapted for “CommonsCasting”, focused on social systems and legal engineering for distributed governance schemes. Key deliverables include a new CVE resource, an opportunity catalog, a technical report, and creative outputs.