“AI companies are asking governments to collaborate internationally to create a level playing field, requiring all companies to prove their powerful AI systems are not dangerous before deployment.”