Policy / Project Profile
International Scientific Exchange on AI Safety
Singapore-hosted scientific exchange mechanism for AI safety: it produced the Singapore Consensus in 2025 and reconvened in May 2026 to update research priorities.
- Category
- International Collaboration
- Published / Updated
- 2025-04
- Issuing body
- IMDA / Singapore AI Safety Institute
- Parliament links
- 1 records
Strategic Context
1 parliamentary record linked.
Detailed Notes
The International Scientific Exchange on AI Safety (ISE) is Singapore’s mechanism for advancing international AI safety cooperation through a scientific-convening format, led by IMDA / the Singapore AI Safety Institute.
ISE 2025 was held as part of the Singapore Conference on AI (SCAI), with the Singapore Consensus on Global AI Safety Research Priorities as its key outcome. ISE 2026, held in Singapore from 17 to 19 May 2026, reconvened experts to update the Singapore Consensus, identify new research priorities, and advance global collaboration in light of the 2026 International AI Safety Report, agent deployments, AI misuse, and capability advances.
The design is deliberately a mix of scientists, government, and industry rather than a purely diplomatic summit. Singapore uses ISE to turn a highly political AI safety topic into a technical research agenda: first define verifiable, collaborative research problems, then let governance frameworks and international mechanisms carry them forward.
Resources
More in This Category
2024-05
Seoul AI Safety Commitment
Ministry of Foreign Affairs (MFA)
2023-11
Bletchley Declaration on AI Safety
Ministry of Foreign Affairs (MFA)
2020
Global Partnership on AI (GPAI)
SNDGO / Ministry of Foreign Affairs (MFA)
2025-05
Singapore Consensus on Global AI Safety Research Priorities
Singapore Conference on AI (SCAI) / Singapore AI Safety Institute