King Stresses Importance of U.S.–China Engagement on AI’s “Shared Risks”

WASHINGTON, D.C. - Ahead of the White House's historic State Visit with President Xi Jinping of the People's Republic of China, U.S. Senator Angus King (I-ME) is underlining the crucial importance of productive engagement by Treasury Secretary Scott Bessent with the Chinese Government on the "shared risks" posed by artificial intelligence (AI). In a letter to Bessent, King acknowledges Bessent's recent public comments on his willingness to discuss critical AI risks with China and states that while the White House should continue to promote American leadership in the artificial intelligence space, continued conversations and dialogue with the Chinese government are critical to AI safety strategy in defense of all humankind.

Senator King wrote, "As you are aware, the American people are increasingly concerned that rapid AI development poses risks that simply cannot be ignored. When an Anthropic employee responsible for ensuring that its models work as intended recently wrote that "[w]e really do earnestly believe AI could kill all humans," many Americans for the first time reckoned with the potentially catastrophic implications of uncontrolled AI development. While thoughtful domestic regulation of the frontier AI labs that are developing the most advanced models is one important aspect of ensuring the safety of the American people, earnest and sustained dialogue with the Chinese government is another critical element of any serious and durable AI safety strategy."

"Chinese and American companies will continue to compete fiercely in the global market for AI products and services, and the U.S. government should continue to work closely with American AI companies to prevent the theft of intellectual property and ensure American leadership in this space," Senator King continued. "But we also cannot lock ourselves into a race to the bottom with China on AI safety given the stakes."

"According to an expert on China's AI policy, the Cyberspace Administration of China has within the last year published an updated "AI Safety & Governance Framework" and created an "AI Safety/Security Working Group" called Working Group 9 to develop policies on frontier AI risks. To ensure that AI will be developed in a safe and sane manner, we should and must work with the Chinese government to create opportunities for American technical experts to exchange proposals with their Chinese counterparts," Senator King concluded.

As former Co-Chair of the Cyberspace Solarium Commission (CSC) and current member of the Senate Armed Services Committee (SASC) and Senate Select Intelligence Committee (SSCI), Senator King is recognized as one of Congress' leading experts on cyber defense, cyber safety, and as a strong advocate for a forward-thinking cyber strategy. Since it officially launched in April 2019, dozens of CSC recommendations have been enacted into law, including the creation of a National Cyber Director.

The full text of the letter can be found here and below.

+++

Dear Secretary Bessent:

Ahead of your meetings this weekend with a delegation from the People's Republic of China led by Vice Premier He and President Trump's meetings with President Xi next week, I write to express my appreciation for your stated willingness to engage with the Chinese government on the "shared risks" posed by Artificial Intelligence ("AI") to both of our nations.

As you are aware, the American people are increasingly concerned that rapid AI development poses risks that simply cannot be ignored. When an Anthropic employee responsible for ensuring that its models work as intended recently wrote that "[w]e really do earnestly believe AI could kill all humans," many Americans for the first time reckoned with the potentially catastrophic implications of uncontrolled AI development. While thoughtful domestic regulation of the frontier AI labs that are developing the most advanced models is one important aspect of ensuring the safety of the American people, earnest and sustained dialogue with the Chinese government is another critical element of any serious and durable AI safety strategy.

Chinese and American companies will continue to compete fiercely in the global market for AI products and services, and the U.S. government should continue to work closely with American AI companies to prevent the theft of intellectual property and ensure American leadership in this space. But we also cannot lock ourselves into a race to the bottom with China on AI safety given the stakes. And there is good reason to expect that the Chinese government is thinking seriously about these issues. According to an American expert on China's AI policy, the Cyberspace Administration of China has within the last year published an updated "AI Safety & Governance Framework" and created an "AI Safety/Security Working Group" called Working Group 9 to develop policies on frontier AI risks.^[1] To ensure that AI will be developed in a safe and sane manner, we should and must work with the Chinese government to create opportunities for American AI safety technical experts to exchange proposals with their Chinese counterparts.

Thank you again for your work to engage productively with the Chinese government on AI safety. I would appreciate staying in touch with you on these issues as Congress continues to debate how to appropriately respond to calls for federal regulation of frontier AI companies.

Sincerely,

###