Applications are now open for the AI Safety Policy Entrepreneurship Fellowship, a programme aimed at people developing practical solutions to emerging challenges surrounding artificial intelligence, including AI safety, security and governance.
The fellowship is designed for participants who want to take an idea beyond research and turn it into a policy proposal capable of influencing governments, regulators and other decision-makers.
Rather than focusing solely on academic research, the programme places emphasis on policy implementation. Fellows will receive guidance on developing their ideas, identifying opportunities for policy change, building partnerships and creating realistic strategies for getting their recommendations adopted.
Fellowship Benefits
Successful applicants will receive a $5,000 stipend for their participation in the programme.
There is also a $1,000 additional merit award for the participant who produces the strongest and most actionable policy memo and advocacy strategy.
Other benefits include:
- Financial support for selected programme-related travel
- Training in policy entrepreneurship
- Professional assistance with editing written work
- Monthly individual mentoring sessions with FAS staff
- Opportunities to connect with specialists and relevant stakeholders
- Invitations to selected FAS events and policy discussions
What Participants Will Work On
Each fellow will select a specific challenge related to frontier AI safety and develop a policy memo proposing a practical response.
The programme will help participants understand how their proposals could move from an idea to actual implementation. This may involve working through legislation, regulation, technical standards, government procurement or other policy mechanisms.
Fellows will participate in training sessions and reading groups while learning how to identify policy opportunities, communicate with decision-makers and build coalitions around their proposals.
The programme also includes a retreat in California and a final capstone event in Washington, DC.
Key Areas of Interest
Applicants can propose work covering a range of emerging AI policy and safety issues, including:
- AI safety research
- AI governance
- AI risk management
- Transparency and accountability
- Cybersecurity
- AI capability thresholds
- Risk classification systems
- Protection of critical infrastructure
- Compute governance
- AI model proliferation
- Management of catastrophic AI risks
Who Is the Fellowship For?
The programme is open to people from a variety of professional and academic backgrounds.
It is particularly intended for early- and mid-career professionals who have expertise in a relevant field but may not have extensive previous experience working directly in public policy.
Potential applicants may come from areas such as:
- Artificial intelligence research
- Academia
- Industry
- Law
- Cybersecurity
- National security
- Civil society
- Think tanks
- Other related technical or policy fields
Applicants should ideally have a clear understanding of an AI-related problem and a practical idea for how governments or other institutions could respond to it.
Programme Details
The fellowship runs from September 30, 2026, through February 28, 2027.
It is a part-time hybrid programme requiring approximately five hours per week, alongside twice-monthly group meetings and other required activities.
The California retreat is currently scheduled for November 4–7, 2026, while the Washington, DC capstone is expected to take place during the week of February 22, 2027.
Application Deadline
Interested candidates have until September 7, 2026, at 11:59 PM Eastern Time to submit their applications.
Those interested should review the eligibility requirements and application instructions on the official FAS fellowship website before applying.
