The Singapore Consensus on Global AI Safety Research Priorities

Bengio, Yoshua, Maharaj, Tegan, Ong, Luke, Russell, Stuart, Song, Dawn, Tegmark, Max, Xue, Lan, Zhang, Ya-Qin, Casper, Stephen, Lee, Wan Sie, Mindermann, Sören, Wilfred, Vanessa, Balachandran, Vidhisha, Barez, Fazl, Belinsky, Michael, Bello, Imane, Bourgon, Malo, Brakel, Mark, Campos, Siméon, Cass-Beggs, Duncan, Chen, Jiahao, Chowdhury, Rumman, Seah, Kuan Chua, Clune, Jeff, Dai, Juntao, Delaborde, Agnes, Dziri, Nouha, Eiras, Francisco, Engels, Joshua, Fan, Jinyu, Gleave, Adam, Goodman, Noah, Heide, Fynn, Heidecke, Johannes, Hendrycks, Dan, Hodes, Cyrus, Hsiang, Bryan Low Kian, Huang, Minlie, Jawhar, Sami, Jingyu, Wang, Kalai, Adam Tauman, Kamphuis, Meindert, Kankanhalli, Mohan, Kantamneni, Subhash, Kirk, Mathias Bonde, Kwa, Thomas, Ladish, Jeffrey, Lam, Kwok-Yan, Sie, Wan Lee, Lee, Taewhi, Li, Xiaojian, Liu, Jiajun, Lu, Chaochao, Mai, Yifan, Mallah, Richard, Michael, Julian, Moës, Nick, Möller, Simon, Nam, Kihyuk, Ng, Kwan Yee, Nitzberg, Mark, Nushi, Besmira, hÉigeartaigh, Seán O, Ortega, Alejandro, Peigné, Pierre, Petrie, James, Prud'Homme, Benjamin, Rabbany, Reihaneh, Sanchez-Pi, Nayat, Schwettmann, Sarah, Shlegeris, Buck, Siddiqui, Saad, Sinha, Aradhana, Soto, Martín, Tan, Cheston, Ting, Dong, Tjhi, William, Trager, Robert, Tse, Brian, H., Anthony Tung K., Wilfred, Vanessa, Willes, John, Wong, Denise, Xu, Wei, Xu, Rongwu, Zeng, Yi, Zhang, HongJiang, Žikelić, Djordje

–arXiv.org Artificial Intelligence 

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).

arXiv.org Artificial Intelligence

Jul-2-2025

arXiv.org    PDF

Add feedback