OpenAI Awards $2M in Grants to Shape AI Policy Globally

On Aug. 17, 2026, OpenAI announced it is awarding grants to 14 independent organizations across the United States, European Union, Brazil, Singapore and South Korea to study the economic and societal consequences of rapidly advancing AI. The total package: $1 million in cash funding plus up to $1 million in API credits, spread across projects ranging from workforce disruption modeling to AI safety frameworks and democratic accountability tools.

The grant program is a direct follow-through on a commitment OpenAI made in April 2026, when it published a policy paper titled Industrial Policy for the Intelligence Age: Ideas to Keep People First. That document argued that incremental policy adjustments would not be sufficient as AI capabilities accelerate toward superintelligence, and called on independent researchers and governments to develop bolder responses. OpenAI received more than 400 proposals before selecting the 14 recipients.

“These choices should not be made by technology companies alone. In democratic societies, they should be shaped through democratic institutions and public debate,” OpenAI wrote in the announcement.

The funded projects divide into two broad categories. The first cluster focuses on building a more open economy: the American Enterprise Institute and the Urban Institute are jointly developing low-, moderate- and high-disruption scenarios for AI’s labor market effects, tied to observable indicators and corresponding policy playbooks. The Progressive Policy Institute will prototype a portable “livelihood insurance” system triggered by broad occupation-level disruption — a model designed to cover workers regardless of whether they hold traditional jobs or gig arrangements. The Tax Foundation will model how AI adoption could shift the composition of public revenues as the balance between labor income, corporate profits and capital gains changes. The European Centre for International Political Economy will examine a proposed “Right to AI” — a framework for ensuring workers and citizens can access AI tools on equitable terms, comparing employee ownership schemes, citizen investment funds and similar models.

The second cluster addresses societal resilience. The Institute for Security and Technology will build a measurement and coordination framework for recursively self-improving AI systems — defining what uncontrolled self-modification looks like, how to classify incidents, and what escalation pathways should connect industry and government. Other projects in this category address biorisks, national security and democratic accountability, with participating institutions including Nanyang Technological University in Singapore and Yonsei University in South Korea.

Brazilian institutions are also represented. The Instituto de Matemática Pura e Aplicada will study what research organizations actually need — compute, engineering support, governance structures — to translate AI access into meaningful scientific output. A major São Paulo public hospital, Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo, will conduct a controlled pre-deployment evaluation of an AI-assisted clinical information system for Brazil’s public health network, using only synthetic or de-identified data.

How This Compares to What Other Labs Are Doing

OpenAI is not the first frontier lab to fund external AI policy research, but its approach has some distinctive features. Anthropic launched its own Economic Futures Program offering $10,000 to $50,000 per grant to academic researchers studying AI’s job-market effects — smaller individual awards focused primarily on academic study. Google’s charitable arm, Google.org, established a $20 million Digital Futures Project in 2023 to fund think tanks and universities on AI policy, a larger pool than OpenAI’s but without the explicit linkage to the funder’s own published policy agenda. OpenAI’s grants are openly framed as an invitation for recipients to test, challenge and build on ideas that OpenAI itself proposed — an unusual posture that blurs the line between philanthropic grant-making and policy advocacy.

That tension has attracted scrutiny. Critics, including researchers at Tech Policy Press, have pointed out that the companies generating the most economic value from AI are simultaneously the most active participants in designing the rules that govern it. Whether OpenAI’s grant-making represents genuine democratic participation or a more subtle form of agenda-setting is a question the funded researchers themselves will be positioned to weigh in on — if they choose to.

For context on the scale of deployment at stake: OpenAI says ChatGPT is now used by 1 billion people worldwide, and claims it has more adult users both under and over the age of 30 than any other frontier lab.

Why This Research Matters If You’re Starting Your Career Now

The policy questions these 14 projects are studying are not abstract for students and recent graduates. Evidence already shows concentrated disruption in specific entry-level roles: writing jobs on at least one major online labor platform declined roughly 2% within four months of ChatGPT’s launch, and freelance writing and translation demand dropped sharply in the same period. At the same time, demand for machine learning programming rose around 24%, and AI-skilled professionals are earning roughly 28% more on average than peers without those skills. The World Economic Forum projects AI will create 170 million new roles globally by 2030, but the distribution of those roles — and who has access to the tools and training to fill them — is far from settled.

The AEI-Urban commission’s disruption scenarios will help policymakers understand which labor market warning signs to watch for and what responses to have ready. The PPI’s livelihood insurance prototype is aimed squarely at workers who move between gig and traditional employment — a pattern common among recent graduates. And the “Right to AI” framework being developed by ECIPE could eventually influence whether access to AI tools in the workplace is treated as a standard employee entitlement or a competitive advantage reserved for those who can afford it.

None of these projects will produce final policy by the time this year’s graduates enter the workforce. But the frameworks, datasets and prototypes they generate are the raw material from which actual legislation and regulation gets built. Paying attention to what they find — and who funds the research shaping AI governance — is part of navigating a job market that is changing faster than most policy can keep up with.

Source: OpenAI