The Singapore AI Safety Hub is recruiting researchers from around the world for a fully funded, three-month residential fellowship that places them at the frontier of one of the most consequential fields in technology. The Singapore AI Safety Fellowship is built for researchers who want their work to land somewhere beyond a journal, in policy rooms, governance frameworks, and the real-world decisions being made about how frontier AI develops.
About the Singapore AI Safety Fellowship Program
Administered by the Singapore AI Safety Hub (SASH) and advised by Ryan Kidd, Co-Executive Director of the MATS Program, the fellowship runs from September 21 to December 4, 2026, based at SASH’s hub in the Chinatown district of Singapore. The program’s core architecture connects fellows from Eastern and Western institutions, treating cross-cultural collaboration as a research priority in its own right rather than a side benefit.
Read Also: Anthropic AI Safety Fellowship 2026 – Apply Now
Funding Size
- Monthly stipend: SGD 5,000 for the full fellowship period
- Full accommodation in Singapore provided
- Round-trip international travel covered
- Up to USD 30,000 in computing resources for projects requiring significant computational capacity
- No application fee
Who Can Apply
- Emerging researchers with a demonstrable track record in technical research
- Candidates with a genuine interest in AI safety, alignment, governance, or responsible AI development
- Researchers comfortable working across cultures, disciplines, and institutional backgrounds
- Applicants must be available to relocate to Singapore on a full-time basis for the entire three-month period
- Visa arrangements are handled by the fellowship for selected international candidates
Geographic Eligibility
- Open globally
- Fellows must be physically present in Singapore for the full fellowship duration
Sector or Thematic Focus
- Technical AI safety and alignment research
- Agent governance and loss-of-control risk
- Frontier AI systems security and robustness
- AI governance, policy translation, and regulatory frameworks
- International coordination on AI safety across Eastern and Western institutions
Application Process
Applications are submitted online through the official SASH portal. The process begins with a written application of approximately one hour covering research intentions, technical background, and AI safety perspectives. Shortlisted candidates complete standardized assessments followed by a mentor-specific work task and interview. Applicants rank preferred research areas and mentors during the application; mentors independently rank shortlisted candidates and the program uses both to make placement decisions.
Required Materials
- Completed online application form covering logistics, research intentions, and fellowship assessment questions
- One technical project, paper, or work sample demonstrating contribution to or interest in AI safety research
- Details of two references
Key Dates
- Application deadline: 10 July 2026
- Online info session (English): 1 July 2026, 8:00 p.m. Singapore time
- Online info session (Mandarin): 2 July 2026, 5:00 p.m. Singapore time
- Fellowship period: 21 September to 4 December 2026
Read Also: The Cambridge Boston Alignment Initiative (CBAI) Fellowship 2026 – Apply Now
Selection Notes
- Projects may be completed individually or in small teams depending on mentor preference and research question
- Expected output is typically a research paper, though alternative formats are accepted where better suited to the project
- The fellowship prioritizes research with a clear pathway to real-world AI safety practice, not only academic publication