SafetyTrustworthinessSecuritycs.AI
The Singapore Consensus on Global AI Safety Research Priorities
Yoshua Bengio, Tegan Maharaj, Luke Ong, Stuart Russell, Dawn Song, Max Tegmark, Lan Xue, Ya-Qin Zhang, Stephen Casper, Wan Sie Lee, Sören Mindermann, Vanessa Wilfred, Vidhisha Balachandran, Fazl Barez, Michael Belinsky, Imane Bello, Malo Bourgon, Mark Brakel, Siméon Campos, Duncan Cass-Beggs, Jiahao Chen, Rumman Chowdhury, Kuan Chua Seah, Jeff Clune, Juntao Dai, Agnes Delaborde, Nouha Dziri, Francisco Eiras, Joshua Engels, Jinyu Fan, Adam Gleave, Noah Goodman, Fynn Heide, Johannes Heidecke, Dan Hendrycks, Cyrus Hodes, Bryan Low Kian Hsiang, Minlie Huang, Sami Jawhar, Wang Jingyu, Adam Tauman Kalai, Meindert Kamphuis, Mohan Kankanhalli, Subhash Kantamneni, Mathias Bonde Kirk, Thomas Kwa, Jeffrey Ladish, Kwok-Yan Lam, Wan Lee Sie, Taewhi Lee, Xiaojian Li, Jiajun Liu, Chaochao Lu, Yifan Mai, Richard Mallah, Julian Michael, Nick Moës, Simon Möller, Kihyuk Nam, Kwan Yee Ng, Mark Nitzberg, Besmira Nushi, Seán O hÉigeartaigh, Alejandro Ortega, Pierre Peigné, James Petrie, Benjamin Prud'Homme, Reihaneh Rabbany, Nayat Sanchez-Pi, Sarah Schwettmann, Buck Shlegeris, Saad Siddiqui, Aradhana Sinha, Martín Soto, Cheston Tan, Dong Ting, William Tjhi, Robert Trager, Brian Tse, Anthony Tung K. H., Vanessa Wilfred, John Willes, Denise Wong, Wei Xu, Rongwu Xu, Yi Zeng, HongJiang Zhang, Djordje Žikelić
June 25, 2025
Robotics
19 citations
1 influential
Key Takeaways
-
AI safety requires a three-layered approach: The report identifies that responsible AI development needs simultaneous focus on three areas—building trustworthy systems from the start (Development), rigorously testing them for risks (Assessment), and maintaining oversight after they're deployed (Control)—rather than relying on any single solution.
-
Global scientific consensus is emerging on safety priorities: The Singapore Conference brought together AI researchers across different countries and backgrounds to identify shared research priorities, building on work supported by 33 governments, indicating that AI safety is becoming a coordinated international concern rather than isolated efforts.
-
Building trust enables both safety and innovation: The report emphasizes that creating a "trusted ecosystem" for AI allows people to confidently adopt these technologies while giving companies space to innovate, suggesting that responsible AI development and business progress are complementary rather than opposing goals.
-
AI safety research is now a recognized scientific field: By synthesizing research priorities across multiple domains and geographies, the report treats AI safety as a legitimate area requiring dedicated scientific investigation, similar to other established fields like cybersecurity or aviation safety.
Abstract
Rapidly improving AI capabilities and autonomy hold significant promise of transformation, but are also driving vigorous debate on how to ensure that AI is safe, i.e., trustworthy, reliable, and secure. Building a trusted ecosystem is therefore essential -- it helps people embrace AI with confidence and gives maximal space for innovation while avoiding backlash. The "2025 Singapore Conference on AI (SCAI): International Scientific Exchange on AI Safety" aimed to support research in this space by bringing together AI scientists across geographies to identify and synthesise research priorities in AI safety. This resulting report builds on the International AI Safety Report chaired by Yoshua Bengio and backed by 33 governments. By adopting a defence-in-depth model, this report organises AI safety research domains into three types: challenges with creating trustworthy AI systems (Development), challenges with evaluating their risks (Assessment), and challenges with monitoring and intervening after deployment (Control).
Final report from the "2025 Singapore Conference on AI (SCAI)" held April 26: https://www.scai.gov.sg/2025/scai2025-report
Paper Info
arXiv ID2506.20702
PublishedJune 25, 2025
RevisedJune 30, 2025
Citations
19
Influential
1
Relevance95%
VenueRobotics
TypeJournalArticle
Categorycs.AI
Responsible AI Topics
SafetyTrustworthinessSecurity
Links
Citation data from Semantic Scholar