Up to $24,000 for three months sounds bold, but the core offer here is time, not just cash. IAPS frames the program for people who can think clearly about frontier AI and its consequences for national security, geopolitics and how governments should respond. It is fully funded and explicitly does not require a technical degree or prior policy experience; what matters is sound judgment and aptitude for grasping what powerful AI could mean for society and the state. Strong research and writing help, prior government or policy work is a plus, yet neither is a prerequisite. Think of it as reading the position rather than memorizing openings in chess — the test is judgment, not recall. The fellowship runs February 22 to May 14, 2027, and applications close on September 27, 2026 through an online form.
Behind it stands the Institute for AI Policy and Strategy, IAPS, a nonpartisan think tank based in the United States. It describes itself as producing practical policy research that stretches from today's most capable models to possible general intelligence and superintelligence. The agenda is not built around writing for its own sake but around responding to priority policy windows that governments actually face. Registered as a U.S. 501(c)(3) nonprofit, its research pillars include national security, verification, international governance and compute policy — from location verification for AI chips to hardware-enabled governance and integrity of AI systems. That framing matters because the fellowship does not lock you to a single lab; it immerses you in the field that moves between think tanks, government and industry. For three months you join a live workflow with a mentor rather than a classroom.
There are two distinct entry routes and the distinction is crucial because the program is not hunting for one researcher archetype. The research track pairs you with an IAPS researcher or an external supervisor at a partner organization to scope a question in the first weeks and then take primary ownership of analysis and drafting for a policy audience, aiming for a publishable piece by the end. Example directions include frontier-model risks, security implications and geopolitical responses — the participant carries the argument from open question to evidence-based conclusions with regular feedback on method and framing. The applied track is built for hands-on policy work rather than a single research output. It suits people coming from adjacent fields such as programs, operations, events, grantmaking or government delivery who want to translate existing professional skills into AI policy experience. Here projects can range from drafting briefs and reviewing legislative drafts to running convenings, evaluating grant applications and shaping technical guidance. If your resume never said AI policy researcher, this second path is a genuine door.
The financial package is tiered and transparent. For full-time participation, Fellows receive about $18,000 and Senior Fellows about $24,000 over the three months; those based in person in DC or London receive an additional $3,000. On top of that comes a $500 professional development allowance, health coverage, and funded travel and accommodation for the mandatory two-week kickoff in Washington, DC from February 22 to March 5, 2027. An invitation letter is provided to support a visa for that kickoff, but the individual visa itself remains the fellow's responsibility. After those two weeks the remaining ten weeks can be done remotely, and exceptional candidates may be offered a part-time track at a pro-rated stipend, though slots are limited. In practical terms the package removes living-cost pressure for a focused quarter; the numbers sit in line with comparable U.S. think-tank fellowships and the residential launch costs are not left to the participant.
Geography is where many outside the United States need clarity. You can choose DC, London or remote. The DC track requires the right to work in the United States between February 22 and May 14, 2027 — IAPS says it cannot sponsor visas for this fellowship. The London track is open to those already based in London and carries the same additional $3,000; IAPS maintains a London hub but notes less active programming than in DC. Remote is where IAPS leans on being remote-first: it states it can legally hire fellows in many countries, yet notes administrative or legal exceptions that vary by individual circumstance. If you are watching from Nigeria, Kenya, Ghana, South Africa, India or elsewhere, the guidance is to not assume worldwide eligibility and to email fellowship@iaps.ai to confirm country-specific hiring. I would call this internationally accessible but not unrestricted. Checking before you apply is the safest step.
Selection is staged and competitive for up to 45 places. Expect initial application review by mid-October, then a written task and a short pre-recorded conversation for those who advance, followed by live interviews in November and final offers by early December. IAPS signals that questions probing how you think carry weight — how you frame a problem, weigh evidence and justify a recommendation matters more than listing credentials. Matching follows selection: the team learns your interests, career goals and skill set and pairs you with a mentor whose workstream you join. Past fellows have contributed to landmark IAPS publications, helped organize policy workshops, and collaborated on memos that entered circulation. In DC the in-person cohort works from the office with desks, monitors and daily lunch, plus regular events that continue learning beyond the kickoff; the London office offers workspace with lighter programming. Even remote fellows join weekly work-in-progress sessions and receive direct feedback from staff.
What happens after is framed as a career bridge rather than a certificate. Fellows receive weekly one-on-one coaching from a fellowship coordinator focused on career development, project management and productivity, plus ongoing support from researchers and access to a professional network. The substance spans chip export controls, verification technology and AI-enabled cyber threats — policy at the technical and institutional seam. Alumni have moved into full-time roles in government, industry and leading think tanks such as AI Futures Project, Future of Life Institute, RAND, the Institute for Progress, the Center for Health Security and the Future Society, while others have founded new initiatives or pursued doctorates. In 2026, when AI governance windows open and close quickly, spending a quarter embedded in a live policy stream offers more than a line on a resume — it offers references, a network and a piece of work that can be shown. With the deadline on September 27, 2026, the calendar is tight; the questions that ask for a thinking sample deserve time rather than a rushed draft. The fellowship page is at www.iaps.ai/fellowship and the contact fellowship@iaps.ai remains open for country checks.
AI commentary
"What surprised me most watching this briefing is that the filter is judgment, not a diploma; the program centers the decision table rather than the lab. For me the value is not only the stipend but the chance to work for three months on live policy questions and plug directly into the field's network."
AI assessment
The strongest case for this program is its diagnosis of a talent gap. As AI governance grows more technical, the policy field needs people who combine judgment, writing and institutional know-how; dropping a technical degree requirement opens the door to strong systems thinkers from outside engineering. Weekly coaching and embedded mentorship deliver craft, not just a report, and the fact that alumni have landed at RAND, Future of Life, Institute for Progress and similar places suggests the bridge works. Steelmanned, the field-learning model beats a seminar: a two-week immersion in Washington repairs the network deficit that remote work alone leaves behind.
Limits are equally clear and should be priced into the decision. With up to 45 places, competition is real and a generic motivation letter will not carry; the questions that ask for a thinking sample demand care. Geographically, the DC track requires U.S. work authorization with no visa sponsorship, which automatically pushes many international candidates toward the remote track. Remote hiring is possible in many countries but exceptions vary by individual circumstance and no public list is shared, so confirmation by email is required. The mandatory two-week DC stay adds passport, visa and family logistics, and the part-time option is limited and exceptional.
On verifiability, two points matter. First, the financial commitments — $18,000, $24,000, the additional $3,000 for DC/London and the $500 development allowance — are consistent across the program page and announcement materials, as are health coverage and funded travel and accommodation for the kickoff. Second, IAPS is presented as an independent think tank with a documented 501(c)(3) registration and a focused research agenda, yet the full map of funding sources and advisory ties is only partially visible to the public. That is less a flaw than a transparency expectation: when weighing policy recommendations, readers also read the funding and network. The funnel — screening by mid-October, live interviews in November, finalization by early December — looks coherent, though the rubric for each stage is not fully detailed.
The practical takeaway depends on profile. If you enjoy synthesis, writing and carrying an open question to publication, the research track fits; if you want to translate program, operations, events or government delivery experience into AI policy, the applied track converts faster. If you live outside the United States, confirm country eligibility by email before you tailor the statement, then build the writing sample to show how you frame a problem, weigh evidence and justify a recommendation. Do not leave it to the last week; as September 27, 2026 approaches, a rushed application will obscure the very thing this program measures — clear judgment.
Sources
6 links; no other published story cites them. Stories sharing a link do not confirm each other; a source's origin is not inferred from how often it is cited.
- @youtube.com YouTube — Funke Suyi: 3-Month AI Fellowship Up to $24,000
- @iaps.ai https://www.iaps.ai/fellowship
- @opportunitydesk.org https://opportunitydesk.org/2026/09/08/iaps-ai-policy-fellowship-spring-2027/
- @effectivealtruism.org https://www.effectivealtruism.org/opportunities/recguVVW08MfF6wAD
- @opportunitiesfeed.com https://opportunitiesfeed.com/2026/09/11/iaps-ai-policy-fellowship/
- @iaps.ai https://www.iaps.ai/research/location-verification-for-ai-chips
artificial intelligence · policy · fellowship · iaps · national security · geopolitics · career