Towards Shared AI Compute Infrastructure

Background

Access to advanced computing infrastructure increasingly shapes which economies can meaningfully influence AI development. This project asks: how can access to large-scale AI compute be structured in a more equitable and sustainable way?

To address this question, the research team will develop a policy framework for regionally shared, jointly financed, and transparently governed AI compute infrastructure. To inform the drafting of the policy brief and complement desk research, the team will host two expert convenings: one preparatory expert convening and one session embedded within a major international AI policy forum scheduled for 2026.

Convenings

• An expert discussion focused on models for establishing sustainable AI infrastructure that can be jointly sponsored and used by multiple participating institutions. Insights from this convening will inform the development of the policy brief.
• Flagship session at an international AI policy forum (2026): A public session to present the policy brief to policymakers, industry representatives, and academic experts, and to solicit structured feedback. This session will aim to situate equitable access to compute as a core issue within broader global AI governance discussions.

Written Outputs

• A policy brief proposing a legal and institutional framework for multinational pooling of AI infrastructure, outlining its key elements, and identifying key challenges and mitigations.
• A series of event summaries to capture discussions & inform the final paper.

Use of Funds

• International AI forum engagement: $5,000
• Research, coordination, and writing support: $18,000 (research and organizational support for the team)

Team Expertise and Track Record

Avinash Kothuri is an MPP and MBA graduate from the University of Oxford. His current research work focuses on benefit sharing of AI, and in the past, he has focused on Special Compute Zones in the EU, the use of responsible AI in the government, and the use of AI for increasing civic engagement in policymaking. Before Oxford, he helped build two tech social enterprises in India. He’s part of the Acumen India Fellowship, Dalai Lama Fellowship, and YLAC Policy in Action networks.

Haakon Huynh is a researcher at the AI Collective Institute and Columbia University. His work focuses on the political economy of artificial intelligence, digital public infrastructure, and public trust in AI governance. He has published on AI in India's digital public infrastructure and U.S. state-level AI regulation. Haakon previously worked with the Royal Norwegian Embassy in India and the United Nations in Geneva. He is an M.A. graduate of Columbia University.