AI Policy Wiki
Dashboard

Federation of American Scientists (FAS)

medium confidence · updated 2026-06-06

Washington nonprofit founded in 1945; Day One Project incubates AI and emerging-tech policy proposals; bridge between academic policy experts and USG.

The Federation of American Scientists (FAS) is a Washington, DC, 501(c)(3) nonprofit founded in 1945 by scientists from the Manhattan Project. Originally focused on nuclear policy and arms control, it now operates as a broad science-and-technology policy organization whose Day One Project solicits and curates near-term executive-branch and congressional policy proposals, including proposals on AI governance, talent, and emerging technology.

FieldValue
Type501(c)(3) nonprofit
HeadquartersWashington, DC
Founded1945 (by Manhattan Project scientists)
CEODaniel Correa (since 2021)
Known forOriginally nuclear policy; now a broad science-and-technology policy shop. Day One Project incubates near-term policy proposals.

History

FAS was founded in 1945 by scientists from the Manhattan Project who held that the technology they had created required deliberate governance. For decades its primary work was nuclear policy and arms control, including the Secrecy Project and nuclear stockpile analysis. Under Daniel Correa's leadership since 2021, FAS shifted toward a broader model of science-and-technology policy entrepreneurship anchored by the Day One Project.

Day One Project

The Day One Project solicits concrete executive-branch and congressional policy proposals from subject-matter experts, curates them, and pitches them to incoming administrations and Congress. It is modeled on the "day-one memo" tradition of presidential transitions. Its proposals span AI governance and national security, AI talent and workforce, semiconductor policy, biosecurity, energy and climate, and public-health data infrastructure.

Day One Project proposals have been cited or adopted in federal AI strategy documents, including aspects of Executive Order 14110 — Safe, Secure, and Trustworthy AI-era implementation and NIST AI work.

Activities and focus

FAS functions primarily as a producer of policy proposals rather than a large-scale research organization on the model of RAND Corporation or Brookings; its work centers on converting expert ideas into policy memos. Its strongest AI focus is AI talent and workforce policy, including proposals to expand federal technical hiring, expand fellowships, and strengthen AI expertise in agencies, a function adjacent to that of Horizon. FAS also publishes regular analysis on export controls, compute, and agency AI adoption. Its nuclear-policy heritage, framed around scientist responsibility for dangerous technology, connects its founding mission to contemporary AI safety and catastrophic-risk policy work.

Relative to other Washington organizations, FAS is less hawkish and more domestic-policy-oriented than Special Competitive Studies Project (SCSP) or Center for Strategic and International Studies (CSIS), more proposal-focused and less research-service-focused than RAND Corporation, and more policy-entrepreneurial with shorter publication cycles than Brookings. Its work is complementary to Horizon, which places people while FAS publishes ideas.

Key people

Daniel Correa is CEO and previously served as Principal Assistant Director for AI/IT at OSTP. Day One Project fellows rotate regularly, and the proposal authors include academics, former USG officials, and industry experts.

Funding

FAS is funded by a mix of foundations, including Hewlett, Open Philanthropy, Gates, and MacArthur, along with individual donors and federal grants. The Day One Project has received specific support from foundations backing evidence-based policy development.

Relationships

Notes

No primary FAS source is currently in the wiki; this page is synthesized. Confidence medium pending source ingest.