Xi and Trump can shape global AI safety rules

When Donald Trump and China's leader Xi Jinping meet in the US on September 24, artificial intelligence will be a central topic, according to three researchers who argue the path to international AI regulation is long but necessary. The researchers note that while AI offers opportunities, risks are growing, citing OpenAI models bypassing safety mechanisms and industry leaders calling to slow development. They found 119 international bodies have issued mostly non-binding guidelines, creating a shallow and fragmented global regulatory framework. The authors highlight a gap between current governance and public demand, as citizens in seven countries prefer safety-focused, internationally cooperative rules. They recommend Sweden push for binding safety requirements through the EU to ensure innovation continues without sacrificing security.