OpenAI funds 14 independent projects on AI policy and resilience
OpenAI has awarded grants to 14 projects led by independent organizations that will explore economic opportunity and societal resilience as artificial intelligence advances. The programme provides a total of $1 million in funding and up to $1 million in API credits, following more than 400 responses to its call for proposals.
The selected projects are based in the United States, the European Union, Brazil, Singapore and South Korea. They will run for six months, with results expected in 2027. OpenAI says the grants follow its April 2026 publication, Industrial Policy for the Intelligence Age, and are intended to test, challenge and develop policy ideas beyond the company itself.
Projects address opportunity and resilience
OpenAI frames the programme around two questions: how AI can broaden economic opportunity, and how societies can build resilience as capabilities advance. The company states that ChatGPT is used by 1 billion people globally, but argues that access alone will not determine how broadly AI’s benefits are shared.
Several grants focus on employment, ownership and public finance. The American Enterprise Institute and Urban Institute bipartisan commission will develop scenarios for AI’s effects on employment and skills, linking observable disruption indicators to policy playbooks. The Centre for European Policy Studies will examine how AI productivity gains are distributed across Europe, while the European Centre for International Political Economy will develop a framework for a proposed “Right to AI.”
Other economic projects include a state-level framework for energy generation and transmission responding to data-centre demand, a person-based benefits model covering different forms of work, and analysis of how AI adoption could affect tax revenue sources. Windfall Trust will expand national working groups across several regions and establish an international group focused on economic preparedness and coordination.
From public services to frontier-risk governance
The grants also cover practical AI deployment and risk governance. Instituto de Matemática Pura e Aplicada will study the support, training, governance, compute and local infrastructure that research institutions need to turn AI access into scientific progress. In Brazil, Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo will conduct a controlled pre-deployment evaluation of an AI-enabled clinical information infrastructure prototype for the Unified Health System, using only synthetic or appropriately de-identified data.
The resilience portfolio includes the Institute for Security and Technology’s framework for measuring and responding to uncontrolled recursively self-improving AI systems. Nuclear Threat Initiative | bio will assess legal feasibility and design an architecture for cross-border sharing of AIxBio capabilities, risks and mitigations. Nanyang Technological University will build privacy-preserving LLM agents for auditable government policy simulation, while Yonsei University will test human-validated AI methods for assessing legislative oversight.
In the context of OpenAI public policy priorities for AI safety, OpenAI’s new grants extend policy engagement from stated priorities toward independently produced research, prototypes, datasets and implementation frameworks. The programme is designed to give governments and public institutions material they can adapt to local circumstances rather than a single prescribed model.
Business implication
For businesses, the projects underline that AI planning increasingly intersects with workforce policy, energy capacity, data protection, public-service workflows and governance. Organisations deploying AI should track the evidence these initiatives produce and assess how emerging policy models could affect their operating assumptions, controls and investment decisions.