Singapore AI Safety Fellowship 2026 (Fully Funded). Apply for fully funded scholarships here. The Singapore AI Safety Fellowship 2026 is a fully funded, three-month research fellowship for emerging AI safety researchers. It’s run by the Singapore AI Safety Hub (SASH), a Singapore-based organization backed by government agencies and international AI safety partners.
Unlike a traditional degree scholarship, this is a residential research placement, not a university program. Fellows relocate to Singapore, work under an experienced mentor, and produce a research output over the three-month period.
Singapore AI Safety Fellowship 2026 Details:
| Offered by: | SASH |
| Scholarship coverage: | Fully Funded |
| Eligible nationality: | All Nationalities |
| Award country: | Singapore |
| Last date: | 10 July 2026 |
Financial Benefits
Fellows don’t need an education loan or personal savings to take part. SASH covers stipend, housing, travel, and even research compute costs, making this a genuinely fully funded opportunity.
| Benefit | Amount / Coverage |
|---|---|
| Monthly Stipend | SGD 5,000 |
| Housing | Fully covered for the 3-month duration |
| Travel | Round-trip travel to and from Singapore covered |
| Research Compute | Up to USD 30,000 per project, for projects that require it |
| Workspace | Central office space at SASH’s hub in Chinatown, Singapore |
Research Focus Areas
Instead of standard academic majors, fellows choose a research track aligned with SASH’s three pillars. Each track pairs fellows with mentors working on that specific problem area.
Verification Mechanisms: Building tools and protocols that prove whether AI systems meet safety claims across training, evaluation, and deployment stages.
Frontier Safety: Developing safety evaluations that catch risks like loss of control or harmful manipulation before frontier models reach public or commercial use.
Agent Governance: Creating mechanisms to identify, supervise, and constrain autonomous AI agents across different regions and use cases.
These three pillars aren’t isolated boxes. Many fellows end up blending technical depth with a policy angle, since SASH explicitly wants research that can inform real-world governance frameworks rather than sitting purely in academic journals.
Confirmed mentors work across a mix of technical and policy-facing roles, including researchers focused on mechanistic interpretability, scalable oversight, and AI control, alongside advisors with backgrounds in international AI governance cooperation.
Host Institution & Mentor Network
Because this is a fellowship rather than a multi-university degree, there’s no single “top universities” list. Instead, fellows work from SASH’s Singapore hub while being mentored by researchers from partner institutions worldwide.
| Institution / Affiliation | Country | Role in Fellowship |
|---|---|---|
| National University of Singapore (NUS) | Singapore | Multiple confirmed mentors, including the Director of the NUS AI Institute |
| Tsinghua University | China | Advisor and mentor contributions from faculty in AI governance research |
| MATS Program / Oxford Martin AI Governance Initiative | USA / UK | Advisory input and mentor affiliations |
Additional mentors are announced closer to the fellowship’s start date, and fellows rank their preferred research area and mentor during the application process.
Advisors to the program include senior figures such as a Chair Professor at Tsinghua University and a Co-Executive Director of the MATS Program, a well-regarded AI safety training initiative. This mix of academic and applied-research backing is part of why SASH frames the fellowship as bridging technical research and real-world policy influence, rather than purely producing academic output.
Eligibility Criteria
SASH is explicit that this fellowship targets researchers, not first-time students. There’s no minimum degree requirement stated, but a strong technical track record is essential.
- A demonstrated track record of strong technical research, such as published work, code, or projects in machine learning or a related field
- Genuine interest in steering AI development toward safe, beneficial outcomes
- Comfort working across cultures and disciplines in an international team
- Full-time availability to relocate to Singapore for the entire three-month duration
There’s no published nationality restriction, and the program is explicitly framed as international, bringing together researchers from both Eastern and Western research communities.
SASH also notes that the fellowship suits people at different career stages, from PhD students and postdocs to independent researchers and industry practitioners, as long as the technical track record is strong. What matters more than a specific title or credential is evidence you can do rigorous, self-directed research without heavy day-to-day supervision.
Applicants juggling other commitments shouldn’t necessarily rule themselves out. SASH’s own FAQ notes that small side responsibilities are usually workable, provided they’re disclosed upfront rather than discovered mid-fellowship.
Required Documents
| Document | Notes |
|---|---|
| Completed online application form | Takes roughly one hour to complete |
| Technical work sample | A project, paper, or codebase showing your AI safety-relevant work |
| CV / resume | Updated with recent research or technical experience |
| Two references | Academic or professional contacts who can speak to your technical ability |
| Mentor and research-area preferences | Ranked as part of the application form |
Application Process
Step 1: Review the fellowship’s three research pillars and decide which area fits your background best.
Step 2: Prepare your strongest technical work sample, whether that’s a paper, an open-source project, or a research write-up.
Step 3: Update your CV to highlight technical research experience relevant to AI safety.
Step 4: Line up two references who can speak specifically to your technical research capability.
Step 5: Complete the online application form, which takes about an hour and includes ranking your preferred mentors.
Step 6: If shortlisted, complete a standardized technical assessment.
Step 7: Complete a mentor-specific work task and/or interview as the final evaluation stage.
This staged process, application, then standardized assessment, then a mentor-specific task or interview, means the strongest applicants aren’t necessarily those with the most polished paperwork. Real technical ability tends to surface at the assessment and task stages, so it’s worth treating those as seriously as the initial form.
Step 8: Wait for mentor matching, which combines your preferences with mentors’ own rankings of candidates.
Step 9: Once accepted, work with SASH on visa arrangements, since the organization sponsors this process for selected fellows.
Step 10: Confirm travel and housing details ahead of the program’s start on 21 September 2026.
Visa & Work Pass Guide
This is where the fellowship differs most from a typical study-abroad scholarship. Since it’s a paid research placement rather than a degree program, fellows don’t apply for a student visa.
SASH states plainly that it arranges visas for accepted fellows, but the organization hasn’t published the specific pass category, processing timeline, or biometric requirements on its public program page. Applicants should ask SASH directly once accepted, since Singapore’s immigration authority, the Ministry of Manpower, sets the exact process depending on the pass type used.
| Item | What’s Confirmed |
|---|---|
| Visa Sponsorship | Arranged by SASH for accepted fellows |
| Visa/Pass Type | Not publicly specified — confirm directly with SASH after acceptance |
| Processing Time | Not publicly specified |
| Remote Participation | Not permitted; the fellowship is fully in-person |
Health Insurance
SASH’s official fellowship page does not publish specific details on health insurance coverage for fellows. This is a genuine gap in publicly available information rather than an oversight on our part.
Given that housing, stipend, and travel are all explicitly covered, it’s reasonable to ask SASH directly whether medical insurance is included or needs to be arranged separately before departure. Singapore itself requires all long-term pass holders to carry adequate health coverage, so this is worth confirming early.
Accommodation & Living Costs
Unlike degree-seeking students who often hunt for their own housing, fellows have accommodation fully covered by SASH for the entire three-month program. This removes one of the biggest cost and logistics headaches of relocating abroad.
Fellows work from SASH’s central office space in Chinatown, one of Singapore’s most walkable and well-connected historic districts. Day-to-day living costs beyond what’s covered, such as personal meals out or leisure spending, follow typical Singapore price levels.
- Casual local meals (hawker centers): roughly SGD 5–10 per meal
- Public transport (MRT/bus): roughly SGD 60–100 per month for regular use
- Personal/leisure spending: varies widely, but SGD 300–600 per month covers a comfortable range
Singapore ranks among the more expensive Asian cities for private rentals, but since fellows don’t need to arrange their own housing, that cost pressure simply doesn’t apply here the way it would for a self-funded student or early-career researcher relocating independently.
After the Fellowship
Because this is a three-month research placement rather than a degree, there’s no formal “post-study work permit” attached to it the way there would be after a university program. Career outcomes instead flow from the research output, mentor relationships, and professional network fellows build.
Singapore has a growing AI safety and governance ecosystem, including organizations like Apart Research, FAR.AI, and the Institute for AI Policy and Strategy, several of which are already connected to SASH as collaborators. Strong fellows may find follow-on opportunities with these groups, with mentor institutions, or in AI safety roles at technology companies operating in the region.
Longer-term work in Singapore would typically go through Singapore’s standard employment pass system administered by the Ministry of Manpower, separate from the fellowship itself.
It’s also worth noting that SASH’s co-working space stays open to people already active in AI governance and safety, or seriously preparing to enter the field. Fellows who want to stay connected to Singapore’s ecosystem after their three months end have a natural on-ramp through that community, even without a formal alumni program being publicly detailed.
Frequently Asked Questions
Is the Singapore AI Safety Fellowship really fully funded?
Yes. It covers a monthly stipend, housing, round-trip travel, and up to USD 30,000 in compute per eligible project, with no application fee.
Do I need a PhD to apply?
No formal degree requirement is published. What matters is a demonstrated track record of strong technical research, regardless of your exact academic stage.
Can I do this fellowship remotely?
No. SASH confirms this is an in-person, residential program, and fellows must relocate to Singapore for the full three months.
What is the application deadline?
The published deadline is 10 July 2026, with the fellowship running from 21 September to 4 December 2026.
Will SASH sponsor my visa?
Yes, SASH states it arranges visas for accepted fellows, though the specific pass type and processing time aren’t detailed publicly.
Can I work part-time or on side projects during the fellowship?
The fellowship itself is full-time, but SASH notes that small side commitments are usually fine if mentioned in your application.
How does mentor matching work?
You rank your preferred research area and mentors on the application form; mentors separately rank the candidates they’d like to work with, and SASH matches based on both.
What happens after the three months end?
There’s no automatic extension or work permit tied to the fellowship. Outcomes depend on your research output, mentor relationships, and any follow-on opportunities within Singapore’s AI safety ecosystem.
Conclusion
The Singapore AI Safety Fellowship 2026 is a genuine, fully funded opportunity for technically strong researchers who want hands-on experience in AI safety and governance, backed by real mentors from NUS, Tsinghua University, and other institutions. It isn’t a university scholarship, so applicants expecting a degree program, student visa, or post-study work permit should recalibrate expectations accordingly.
With the application deadline set for 10 July 2026, interested researchers should apply directly through the official SASH website as soon as possible, and confirm any unpublished details, like visa type or insurance, directly with the organization after acceptance.
Official Resources
| Website Type | Official Website | Latest Application Deadline | Status |
|---|---|---|---|
| Official Fellowship Page | SASH – Singapore AI Safety Fellowship | 10 July 2026 | Applications Open — Closing Soon |
| Official Application Form | SASH Fellowship Application Form | 10 July 2026 | Applications Open |
| Host Organization | Singapore AI Safety Hub (SASH) | N/A | Official |
| Contact SASH Directly | SASH Contact Page | N/A | Official |
| Singapore Immigration / Work Passes | Ministry of Manpower, Singapore | N/A | Official |
Note: The application deadline of 10 July 2026 is only days away at the time of writing. Visa/pass type, insurance details, and remote-participation rules are not fully published by SASH — confirm these directly with the organization before or after applying.