OpenAI is putting money behind the policy ideas it floated earlier this year. The company announced on August 17, 2026 that it is awarding grants to 14 projects led by independent organizations, aimed at researching how AI is likely to reshape economic opportunity and social stability as the technology advances. The grants go to universities, think tanks, and nonprofits across the United States, the European Union, Brazil, Singapore, and South Korea.
The announcement, published on OpenAI’s Global Affairs blog, follows through on a commitment the company made in April, when it published a policy paper called “Industrial Policy for the Intelligence Age.” That paper proposed a slate of ideas — from a public wealth fund to a four-day work week — and promised to fund outside researchers who wanted to test, challenge, or build on them. Four months later, OpenAI says it has picked its recipients from a pool of more than 400 applicants.
Table of Contents
What OpenAI Announced
According to OpenAI, the company will provide a total of $1 million in direct funding plus up to $1 million in API credits, spread across the 14 selected projects. The company has not disclosed how that funding is divided among individual grantees, so it’s not clear whether any project received a disproportionately large or small share.
Each project is expected to run for six months, with results reported sometime in 2027. OpenAI frames the initiative as one part of a broader effort to “widen the circle” of people working on AI policy questions, rather than leaving those questions to technology companies alone.
It’s worth being clear about what this is and isn’t. This is a research-funding announcement, not a new product, regulatory filing, or technical release. The figures involved — the funding total, the applicant count, the list of grantees — come from OpenAI’s own account of the program. No independent organization has yet audited or verified these numbers.
Background: The “Industrial Policy for the Intelligence Age” Paper
To understand why OpenAI is funding this research, it helps to go back to April 6, 2026, when the company published its policy paper “Industrial Policy for the Intelligence Age: Ideas to Keep People First.” That document argued that incremental policy tweaks won’t be enough to manage the transition AI is expected to bring, and it laid out a set of proposals that included a public wealth fund seeded by AI companies, pilot programs for shorter work weeks, investment in electric grid capacity, and a stronger social safety net.
OpenAI described these ideas explicitly as a starting point, not a finished blueprint. Alongside the paper, the company said it would establish “a pilot program of fellowships and focused research grants of up to $100,000 and up to $1 million in API credits” for people working on related ideas, and it opened an inbox for proposals.
The response was substantial. By June 9, 2026, OpenAI said it had received more than 400 submissions and closed the application window to begin reviewing candidates. The 14 grantees named in August are the outcome of that review process.
It’s worth noting that the per-project ceiling mentioned in the April paper — up to $100,000 and $1 million in API credits — describes what an individual project could receive under the pilot framework. The $1 million total funding figure announced in August applies to the entire slate of 14 projects combined, not to each one individually.
The 14 Funded Projects
OpenAI grouped the projects under two broad questions: how AI can broaden economic opportunity, and how societies can build resilience as AI systems become more capable.
Building a More Open Economy
Most of the funded work falls into this category, focusing on labor markets, taxation, ownership, and access.
| Organization | Region | Project Focus |
|---|---|---|
| American Enterprise Institute (with Urban Institute) | United States | Bipartisan commission modeling low-, moderate-, and high-disruption scenarios for AI’s effect on employment and skills |
| Centre for European Policy Studies | European Union | Examining how AI-driven productivity gains are distributed across workers, firms, and regions in Europe versus the US |
| European Centre for International Political Economy | European Union | Developing a “Right to AI” framework comparing employee ownership, citizen investment, and social wealth funds |
| Abundance Institute | United States | Comparing how US states can expand energy generation for data centers while producing local benefits |
| Progressive Policy Institute | United States | Designing a “person-based benefits” system covering retirement, health, leave, and disability across different forms of work |
| Tax Foundation | United States | Analyzing how AI adoption could shift the balance between labor income, corporate profits, and other tax revenue sources |
| Windfall Trust | United States, UK, Canada, EU, Latin America | Scaling national working groups examining fiscal and labor-market questions under different AI scenarios |
| Instituto de Matemática Pura e Aplicada (IMPA) | Brazil | Studying what research institutions need to convert AI access into scientific progress |
| Hospital das Clínicas (University of São Paulo) | Brazil | Evaluating a prototype AI-enabled clinical information system for Brazil’s public health system, using synthetic or de-identified data |
Building a More Resilient Society
A smaller cluster of projects looks at governance, oversight, and safety-related coordination rather than pure economics.
| Organization | Region | Project Focus |
|---|---|---|
| Institute for Security and Technology | United States | Building a measurement and coordination framework for tracking recursively self-improving AI systems |
| Nuclear Threat Initiative (NTI | bio) | International | Testing the legal feasibility of secure, cross-border information-sharing on AI-related biosecurity risks |
| Council on Strategic Risks | United States | Training national-security professionals in frontier AI safety and governance |
| Nanyang Technological University | Singapore | Building an AI-agent framework to simulate household responses to government policy changes |
| Yonsei University | South Korea | Testing whether AI methods can help assess the quality of legislative oversight in South Korea’s National Assembly |
Notably, the Institute for Security and Technology and Nuclear Threat Initiative projects sit closer to AI safety and governance than to economic policy narrowly defined — a reminder that OpenAI’s framing of “resilience” extends beyond labor markets and into questions about how AI systems themselves are monitored and controlled.
OpenAI has named the institutions leading each project but has not published the names of individual lead researchers, nor has it detailed the criteria used to choose 14 winners out of more than 400 applicants.
How This Compares to Anthropic’s Economic Futures Program
OpenAI isn’t the only AI lab funding outside research into AI’s economic effects. Anthropic launched a comparable effort, the Economic Futures Program, in November 2025. That program offers Economic Futures Research Awards of between $10,000 and $50,000 for empirical research on AI’s economic impacts, alongside policy symposia and Anthropic’s own Economic Index, which tracks how people actually use its Claude models across the economy.
Anthropic’s commitment to this space is larger in dollar terms when its full policy push is considered. On June 10, 2026, the company published its Advanced AI Framework and Economic Policy Framework alongside a pledge of $350 million, split between a $200 million Economic Futures Research Fund focused on studying job displacement and a $150 million national fellowship program for early-career Americans.
| OpenAI | Anthropic | |
|---|---|---|
| Program name | Industrial Policy grants (2026) | Economic Futures Program |
| Total commitment (this initiative) | $1M funding + up to $1M in API credits | Award-based, $10K–$50K per grant |
| Broader related pledge | Not applicable | $350M ($200M research fund + $150M fellowships) |
| Number of funded projects | 14 | Ongoing rolling awards |
| Project timeline | 6 months | 6 months (per award) |
Anthropic’s award range reflects an ongoing, rolling grants program rather than a fixed total commitment for this specific initiative.
The two programs aren’t structured the same way, so a direct dollar-for-dollar comparison is misleading. OpenAI’s initiative is a one-time slate of 14 named institutional grantees tied to a specific policy paper, while Anthropic’s is an ongoing awards program combined with a separate, much larger financial pledge covering both safety and economic policy work. No comparable named program from Google DeepMind, Meta AI, or Microsoft turned up in available reporting.
Why It Matters
The significance of this announcement isn’t really about the dollar figures, which are modest by the standards of the AI industry. It’s about what the move signals: a leading AI lab is trying to shape the terms of the policy conversation around AI’s economic disruption before that disruption fully arrives, rather than waiting for governments to set the agenda.
That’s consistent with a broader shift in how frontier AI companies talk about their responsibilities. Where safety commitments once centered narrowly on technical alignment and misuse prevention, OpenAI’s framing here extends into labor markets, tax policy, and institutional resilience — areas traditionally the province of governments and independent policy researchers, not technology vendors.
Whether that shift is welcomed as constructive engagement or viewed skeptically as a company writing its own regulatory narrative depends partly on who’s asked.
Open Questions and Limitations
Several details about the program remain undisclosed. OpenAI has not published the names of individual researchers leading each project, nor has it explained in detail how it narrowed more than 400 applicants down to 14. It’s also unclear whether all the research produced under these grants will be released under an open license — OpenAI’s post suggests most projects will produce public-facing reports, webinars, or datasets, but it stops short of committing to full open access for every project’s underlying data.
There’s also a structural question that critics have raised about OpenAI’s broader policy push, separate from the specific 14 grantees. When OpenAI’s April paper was published, Anton Leicht, a visiting scholar at the Carnegie Endowment for International Peace, described it as “comms work to provide cover for regulatory nihilism” — essentially arguing that big policy ideas can function as a way to project responsibility while a company continues to build and deploy AI systems at speed. Commentary from outlets like Tech Policy Press raised similar concerns, noting that OpenAI’s proposals would deepen the company’s integration into economic and physical infrastructure while asking governments to help fund that expansion.
It’s important to be precise about what this criticism does and doesn’t apply to. These critiques were directed at OpenAI’s original policy paper and the company’s motives more broadly — not at the 14 independent organizations now receiving grants, none of which have publicly commented on their own independence from OpenAI’s goals. Because OpenAI is both the author of the ideas being tested and the funder of the research testing them, the question of how independent that research will be in practice is a reasonable one to watch, even without evidence of any problem in this specific case.
What Happens Next
The 14 projects are just getting underway. With six-month timelines, most should produce initial findings by early-to-mid 2027, and OpenAI says the reported results will feed into “a much broader public debate” about AI policy — though the company doesn’t control what that debate concludes, or whether governments act on any of it.
For now, the concrete, verifiable facts are limited to what’s been announced: 14 named institutions, a combined $1 million in funding plus matching API credits, and a six-month research window. Everything downstream of that — what the researchers find, whether it changes any government’s approach, and whether the “Right to AI,” public wealth funds, or other ideas from OpenAI’s April paper gain real traction — remains to be seen.
Conclusion
OpenAI’s grant announcement is a modest but concrete step that turns an earlier policy paper into funded, if not fully independent by construction, research. The list of 14 grantees — spanning economic think tanks, universities, and security-focused institutes across five regions — gives outside observers something specific to track over the next year, rather than another set of proposals with no follow-through. What’s genuinely established is the funding commitment and the list of recipients. What remains open is how the selection process worked, who’s leading the research day to day, and whether findings due in 2027 will meaningfully shape policy or simply add to an already crowded conversation about AI’s economic future.
FAQ
What did OpenAI announce on August 17, 2026?
OpenAI announced it is funding 14 independent research projects, led by organizations across the US, EU, Brazil, Singapore, and South Korea, to study how AI affects economic opportunity and societal resilience.
How much funding is involved?
OpenAI says it will provide a total of $1 million in direct funding plus up to $1 million in API credits, spread across all 14 projects combined. The company has not disclosed a per-project breakdown.
Which organizations received grants?
Recipients include the American Enterprise Institute (with the Urban Institute), the Centre for European Policy Studies, the European Centre for International Political Economy, the Abundance Institute, the Progressive Policy Institute, the Tax Foundation, Windfall Trust, Brazil’s IMPA and Hospital das Clínicas, the Institute for Security and Technology, the Nuclear Threat Initiative, the Council on Strategic Risks, Nanyang Technological University, and Yonsei University.
How does this relate to OpenAI’s “Industrial Policy for the Intelligence Age” paper?
This grant program fulfills a commitment OpenAI made in that April 2026 paper, which proposed policy ideas like a public wealth fund and pledged to fund outside researchers who wanted to test or build on them.
When will results be available?
Each project runs for six months, with OpenAI saying results will be reported in 2027.
How does this compare to Anthropic’s approach to AI economic policy?
Anthropic runs a comparable Economic Futures Program offering smaller individual research awards ($10,000–$50,000), alongside a separate $350 million pledge announced in June 2026 covering both economic research and a national fellowship program. The two companies’ initiatives differ in structure and aren’t directly comparable dollar-for-dollar.
Are the funded researchers independent of OpenAI?
The 14 organizations are institutionally separate from OpenAI, but none have publicly commented on their independence from the company’s goals. Some outside critics have questioned the broader motives behind OpenAI’s original policy paper, though that criticism was not directed at the specific grantees.