Frontier AI Safety in 2026: Evidence, Risks, and the Road Ahead

​Frontier AI Safety in 2026: Evidence, Risks, and the Road Ahead is hosted by Concordia AI, with the support of Singapore’s Infocomm Media Development Authority (IMDA) and SGInnovate, as part of Singapore AI Research Week 2026.

​Our understanding of frontier AI risks has come into sharper focus. Concerns around cyber and biological misuse, deceptive behaviour, and misalignment from frontier AI systems are being substantiated by a growing body of evaluations and empirical evidence. This session surveys the AI safety landscape in early 2026 and the tools used to monitor and manage these risks.

​The programme will feature focused presentations on:

• ​Latest updates from the International AI Safety Report

• ​​The Frontier AI Risk Management Framework (Shanghai AI Lab)

• ​​The Frontier AI Risk Monitoring Platform (Concordia AI)

​​These presentations will be followed by a moderated panel discussion and conclude with informal networking and refreshments.

​​Whether you are a technical AI safety researcher, an industry professional, or simply interested in understanding how frontier AI systems can be developed and deployed responsibly, this session will offer timely insights and thoughtful discussion.

​RSVP for free if you want to attend. Details on the agenda and panellists will be shared closer to the event.