Frontier AI capabilities and autonomy are advancing rapidly. A growing number of real-world incidents make a trusted AI ecosystem essential to embracing AI with confidence. The 2026 Singapore Consensus is an outcome of the second International Scientific Exchange on AI Safety, bringing together over 100 contributors spanning 13 countries from frontier developers, government safety institutes, academia, and civil society. Building on the 2025 report, it presents a global understanding of technical AI safety research problems of top priority—now with a dedicated focus on societal resilience and on managing the risks of increasingly autonomous AI agents.
The 2026 Singapore Consensus also takes a deep dive into risks stemming from the fast-growing development and deployment of autonomous agents. It proposes managing these risks through ten foundational principles spanning the agentic AI lifecycle. These are categorised into 1. Design and development (least privilege, traceable identity, auditability);2. Testing and deployment (validated deployment, adversarial resilience, multi-agent stability); and 3. Operation and monitoring (runtime assurance, interruptibility, legibility, and human oversight). These ten principles are described in depth in the Companion Report on Agentic Risk Management.
The 2026 Singapore Consensus is the outcome of the International Scientific Exchange on AI Safety (ISE), held in Singapore in May 2026 with support from the Future of Life Institute and Concordia AI. Brian Tse served as a Steering Committee member, Oskar Galeev as a Lead Writer, and Kwan Yee Ng as a writing contributor to the Consensus. For more information, please visit the official website.