From April 17 to 19, 2026, during the fifth session of the International Dialogue on AI Safety (IDAIS), hosted by the Royal Society in the United Kingdom, a group of preeminent scientists and governance experts convened to address the pressing concerns surrounding artificial intelligence. Among them were Turing Award laureates Yao Qizhi, Dean of the School of AI at Tsinghua University, and Yoshua Bengio, alongside Zhang Yaqin, the founding Dean of the Institute for AI Industry Research (AIR) at Tsinghua University, and Professor Stuart Russell from the University of California, Berkeley. Together, they collaboratively drafted and endorsed a pivotal statement.

The signatories cautioned that, given the current trajectory, non-state actors with modest technical prowess could acquire cyberattack capabilities on a national scale within the next year, thereby presenting unprecedented threats to vital infrastructure and national security. Furthermore, the potential for misuse in the biological domain is becoming an increasingly significant concern.

In response to these challenges, the statement urges nations to collectively confront the shared threats posed by AI-driven cyberattacks and the proliferation of technologies that could be misused in biological contexts. It outlines critical priorities, including the protection of essential infrastructure, enhancement of the security measures for cutting-edge AI models, and the establishment of a robust, all-encompassing risk assessment framework.