Washington, DC -- Representatives Gregory W. Meeks, Ranking Member of the House Foreign Affairs Committee, Ami Bera, Ranking Member of the Subcommittee on East Asia and the Pacific, Sara Jacobs, Ranking Member of the Subcommittee on Africa, and twenty-six additional Democrats today sent a letter to Secretary of State Marco Rubio and Treasury Secretary Scott Bessent calling for a sustained and substantive dialogue with China to address AI-related safety risks, and urged the administration to initiate such discussions at the U.S. bilateral engagement with China next week. The letter called on the administration to aim to:
- Establish shared standards and red lines for AI safety
- Develop a standing crisis communications channel to facilitate coordination in the event of an AI-related emergency and to report anomalous events
- Address proliferation of advanced AI capability to third parties
- Facilitate dialogue between American and Chinese AI researchers on AI safety and risks.
A PDF copy of the full letter can be found here. An excerpt is below:
“...Advancements in artificial intelligence have the potential to dramatically improve and transform our way of life. They also present a broad spectrum of national security risks that could be harmful to the American public and our national interests. AI models could be used to create a deadly pathogen, carry out a cyberattack on critical infrastructure, or destabilize societies in other ways. Models trained in one part of the world can be misused in another, with the consequences spilling far beyond a nation’s borders. Last week, researchers from Anthropic and Open AI — America’s top two AI labs — warned the world that the technology is hurtling toward a future that, minus guardrails, could endanger humanity.
“To mitigate the risks of AI at home, we must also address the risks internationally, especially with China given its capacity to develop advanced AI models of its own. Though there is much we disagree on, our nations are also vulnerable to the same AI risks. Yet, right now, the two countries developing the most capable AI systems in the world have no reliable means to inform one another of or jointly address risks. The purpose of a U.S.-China dialogue on AI should be focused and practical: reduce risks of miscalculation and address the dangers that neither country can manage alone.”
The co-signers of the letter include: Representatives Gabe Amo, Yassamin Ansari, Donald S. Beyer Jr., Greg Casar, Joaquin Castro, Jim Costa, Veronica Escobar, Jonathan L. Jackson, Sydney Kamlager-Dove, Ted W. Lieu, Sarah McBride, James P. McGovern, Grace Meng, Kweisi Mfume, Dave Min, Eleanor Holmes Norton, Johnny Olszewski, Jr., Kristen McDonald Rivet, Mary Gay Scanlon, Bradley Scott Schneider, Hillary J. Scholten, Brad Sherman, Greg Stanton, Dina Titus, Debbie Wasserman Schultz, and George Whitesides.