The OpenAI Foundation has announced a $250 million commitment to grants, partnerships, and direct work focused on what it calls "economic futures in the age of AI." The initiative, detailed in a post published today by Divya Siddarth and Wojciech Zaremba, marks the Foundation's most concrete attempt yet to address AI's labor market effects.

The announcement follows the Foundation's broader March pledge to invest at least $1 billion over the coming year across health, jobs, AI safety, and community programs. This $250 million slice is earmarked specifically for understanding and responding to how AI reshapes work and economic security.

Three Pillars, Many Open Questions

The program divides into three areas. The first, "Understanding the Shift," funds independent measurement and forecasting infrastructure. The Foundation argues that existing statistical systems were built for a different era and fail to capture how AI-generated value flows between workers, firms, consumers, governments, and capital owners. It calls for modernized labor market data systems worldwide, including Bureau of Labor Statistics-style capacity in more countries and updated occupational mapping tools like O*NET.

The second pillar, "Supporting the Transition," resources workers and communities through near-term disruption. The Foundation says this could include support during job searches, expanded unemployment and wage-loss insurance, help translating experience into new roles, and pathways into growing sectors. It acknowledges that "traditional retraining programs have mixed evidence" and suggests an AI transition agenda will need to be broader. The Foundation also wants to invest in government capacity to deliver these programs, including using AI itself to accelerate public services.

The third area, "Building Economic Security," ventures into more contested territory. The Foundation says it wants to fund research and pilots exploring revenue mechanisms like shifting taxation from labor to capital, windfall taxes on extraordinary returns, and public or sovereign wealth funds modeled on Norway's Government Pension Fund or Alaska's Permanent Fund. On distribution, it mentions income, capital stakes, public goods, essential services, jobs programs, and access to compute.

Who Is Writing This?

The announcement is co-authored by Divya Siddarth, co-founder and executive director of the Collective Intelligence Project, an organization that has partnered with OpenAI, Anthropic, and government bodies including Taiwan's Digital Ministry and the UK AI Safety Institute on democratic input processes for AI development. Siddarth's background includes positions at Microsoft's Office of the CTO, Microsoft Research India, and Stanford.

Wojciech Zaremba is an OpenAI co-founder who earlier this year joined the OpenAI Foundation as Head of AI Resilience. He had previously led work on OpenAI's robotics, GPT models, and Codex.

The pairing signals the Foundation is drawing on external expertise in governance and collective decision-making, not just internal OpenAI staff.

The Context: $1 Billion, Many Skeptics

The economic futures initiative arrives as the Foundation scales up its philanthropic footprint. In December, it disbursed $40.5 million in grants to 208 U.S. nonprofits through its People-First AI Fund. Critics have questioned whether the Foundation can operate independently from OpenAI's commercial interests, particularly on questions where safety research or policy recommendations might conflict with the for-profit company's product decisions.

The $250 million figure represents a meaningful sum, but it is small relative to the Foundation's assets and the scale of economic disruption AI may cause. The announcement acknowledges deep uncertainty about pace and scale, stating: "We don't need to know exactly how the future will unfold to prepare for it."

In April, OpenAI separately released policy proposals advocating for public wealth funds, portable benefits, and shifting taxes from labor to capital. The Foundation's new initiative appears to fund research and pilots that could inform such proposals.

Measuring What Matters

The most concrete element of the announcement is its focus on measurement infrastructure. The Foundation notes that if AI generates value as digital goods rather than wages, income statistics will miss it. If labor's share of income declines, workers' bargaining power may weaken and GDP may become a worse proxy for welfare. The initiative specifically calls for tracking what people can actually do and access, not just what they earn.

This framing suggests the Foundation views existing economic indicators as inadequate for an AI transition. Whether it can fund alternatives that governments and researchers actually adopt is an open question.

The Foundation says it is inviting public input on what people are experiencing in their work and communities. First initiatives are expected later this year.