Stay informed with free updates
Just sign up for artificial intelligence myFT Digest — delivered straight to your inbox.
The White House’s top science adviser said the United States would work with China on the safety of artificial intelligence systems in the coming months, hinting at rare cooperation between the two countries.
Arati Prabhakar, director of the White House Office of Science and Technology Policy, told the Financial Times that despite the US-China trade tensions over AI, the two countries will work together to reduce its risks and assess its capabilities. Told.
Regarding future cooperation with China on AI, Prabhakar said, “We have taken steps to be involved in that process.” “We have to work [with Beijing]”
Her comments are a clear signal that the two countries plan to work together to protect rapidly evolving technology, even at a time of heightened trade tensions between the two countries.
In November, China signed the UK’s Bletchley Park Agreement on standards for this technology, while US President Joe Biden and Chinese President Xi Jinping discussed cooperation in the field of AI at a summit in California that month. .
“We are at a moment when everyone understands that AI is the most powerful technology…countries are preparing to use it to build a future that reflects their values. Prabhakar, who advises Biden on issues such as AI regulation, said.
“But I think the one thing we can all really agree on is that we want to have a safe and effective technology base,” she added. “So I think it’s a good place for collaboration.”
Earlier this month, the Financial Times reported that major US AI companies, including OpenAI, held secret meetings with Chinese experts to discuss new risks associated with the technology.
There are growing concerns around the world that rapid advances in AI will exacerbate vulnerabilities to cyberattacks and spread misinformation.
Prabhakar said the best options available for assessing the safety of new AI systems are “less than zero” due to the complex and opaque nature of the technology.
“We have shown the world that the tools to evaluate AI models – to understand how effective, safe and reliable they are – are extremely weak today. “I’m starting to understand that,” she said.
The comments come as the United States imposes export controls on chips and equipment to hinder China’s ability to develop advanced technologies such as AI.
Big tech companies in the U.S. and China are also racing to develop products that leverage generative AI, models that can spit out human-like text and materials in seconds.
Countries have chosen to regulate domestic AI groups in different ways. For example, China has emphasized the need for content management and provided detailed guidance on AI development.
By contrast, Biden issued a sweeping executive order on AI last October to address threats, with a focus on national security and consumer privacy.
Prabhakar said that while China and the U.S. may disagree on certain values and regulatory approaches, “there are some things we can agree on,” such as global technical and safety standards for AI software. .
He said the United States has no intention of slowing down AI development and intends to maintain oversight of the technology. But the White House is acting too quickly to regulate AI, jeopardizing America’s competitive advantage in AI, even though no specific legislation on the issue has been passed. is facing criticism.
“In the United States, we understand that we are at a point where our leadership in the world depends on our leadership in AI,” Prabhakar said.
But even U.S. AI companies, she argued, have accepted the need for a clear way to understand and evaluate AI, which is critical to consumer trust and enterprise adoption. “Technology champions within companies are some of the loudest advocates for regulation,” Prabhakar said.
Ryan Hass, director of the Brookings Institution’s China Center, said cooperation on AI is “not about goodwill or charity. It’s a hard calculation.”
He added: “Given the speed at which technology is evolving, there is momentum to do something now. It will be important for both sides to demonstrate early on that the dialogue can yield results. right.”
Additional reporting by Demetri Sevastopulo in Washington