The Federation of American Scientists (FAS) has launched a new AI Safety Policy Entrepreneurship Fellowship aimed at helping professionals turn ideas on artificial intelligence safety and governance into concrete public policies.
The six-month, part-time programme will run from September 30, 2026, to February 28, 2027. Selected fellows will work with the FAS to develop policy proposals addressing emerging risks associated with artificial intelligence and identify practical pathways for putting those proposals into effect.
The programme offers a $5,000 stipend to each fellow, with an additional $1,000 potentially available to the participant whose policy memorandum and advocacy strategy are judged the most precise and actionable.
The fellowship is designed as a hybrid programme and requires an estimated five hours of work per week. Participants will receive mentoring, policy training, editorial support and access to experts and policymakers.
The FAS is inviting applications from a broad range of professional backgrounds. Candidates may come from fields including artificial intelligence, cybersecurity, academia, law, industry, civil society, national security and public policy. Direct experience in government policy is not required.
Projects can focus on areas such as AI safety, cybersecurity, critical infrastructure, transparency, accountability, AI risk assessment, computing governance and the regulation of advanced AI systems.
The programme comes as governments and policymakers around the world face growing pressure to develop rules capable of addressing the risks created by increasingly powerful AI systems while preserving technological innovation.
The fellowship seeks to bridge the gap between research and policymaking by helping participants transform technical or policy ideas into proposals that could be adopted through legislation, regulation, standards, institutional reforms or government procurement policies.
Applications are open until September 7, 2026, at 11:59 p.m. Eastern Time.
For the Federation of American Scientists, the initiative represents an effort to build a new generation of policy entrepreneurs capable of translating AI safety research into practical measures designed to reduce emerging technological and security risks.
