By Jeffrey Dastin
SAN FRANCISCO, Aug 26 (Reuters) - Bill Gates is looking to meet with Chinese President Xi Jinping later this year, eager to propose global efforts to mitigate the growing risks posed by artificial intelligence.
The Microsoft co-founder and billionaire philanthropist told Reuters in an interview that he believes China might agree to restricting potentially dangerous AI model releases if the U.S. took initiative on that first.
"The whole world would go along with that," said Gates, who met with Xi three years ago, when he was the first foreign entrepreneur to be hosted by the Chinese president post-pandemic.
Gates also said he'd like to propose that countries monitor AI's ability to create molecules and carry out biological attacks — something that would not get in the way of any tech industry's development.
Gates' staff are working to finalize a meeting with Xi for his next China trip, his spokesperson said, scheduled tentatively for November.
President Xi "might be surprised by how vehement I am," Gates said. He said China is protecting children from AI companion addiction, but he believes few leaders globally are rising to the occasion to get ahead of a litany of potential harms from AI.
"Yes, in the past, technology never did this. This time is different. This is like an evolutionary milestone that needs incredible policy management by society as a whole."
AI HACKING 'SHOULD BLOW PEOPLE'S MINDS'
Gates struck a more optimistic note about AI in January, but since then he has become concerned that it can conduct attacks while supplanting jobs and hooking users psychologically.
For instance, technology from OpenAI, Anthropic and Meta Platforms recently hacked real-world websites in what were supposed to be isolated cybersecurity evaluations. "These security incidents are shocking, and they should blow people's minds," Gates said.
In a blog post on Wednesday, Gates called for a new international organization to manage AI, drawing on elements from a nuclear inspections regime, international aviation regulations, and ozone layer protection agreements.
To date, U.S. President Donald Trump's administration has exercised some control over the distribution of powerful AI models such as Anthropic's Fable 5, but more regulation has taken place at the state level.
Whereas many tech leaders have, like Meta CEO Mark Zuckerberg, touted the benefits of AI, many have also expressed concern about its perils. Demis Hassabis, chair of Google DeepMind, proposed a body like the Financial Industry Regulatory Authority, or FINRA, that could develop and conduct tests of AI related to national security.