Just over a week ago, I attended a major artificial intelligence conference in Zhongguancun, Beijing’s bustling high-tech district. It was packed with fascinating sessions touching on everything from recursive self-improvement—the idea that models can tweak their own code and advance indefinitely—to humanoid robots.

But what struck me most was the need for cooperation between the US and China. As Stephen Casper, a computer scientist at MIT who spoke via video, noted: ‘AI is a global technology with global benefits, global harms, and a consistent tendency for new capabilities to eventually proliferate.’

The world’s two dominant AI powers are responsible for the most advanced models, and as these models become more powerful and intertwined in our daily lives, the risks of cyberattacks or catastrophic failures grow. The US has largely viewed China’s AI advances as an economic and national security threat, but the conference reinforced the idea that both nations stand to lose if AI is developed too quickly and recklessly.

Lin Yun, a professor at Shanghai Jia Tong University who does excellent work on AI and computer security, expects hackers to gain an advantage over the near term. However, new countermeasures, including novel uses of AI, should tip the balance back toward defense over time. He said that even if international cooperation is complicated by competition, it should remain a priority as shared safety principles can reduce systemic risk.

The most pressing question for both nations is how to balance openness with risk. Open models have become crucial for research and innovation, but as these models advance, ensuring they don’t help hackers identify security vulnerabilities or be wielded as cyber weapons will become more challenging.