Representative Ro Khanna, the top Democrat on the House Select Committee on China, has sent letters to three leading Chinese AI companies and a US intelligence agency as part of a push for a bilateral safety agreement. The letters to DeepSeek, Alibaba, and Moonshot AI request documentation on their work toward superintelligence and recursive self-improvement, along with details on safeguards and kill switches. He also asked the Office of the Director of National Intelligence to assess how well the US could respond if an AI lab lost control of its systems, and to evaluate China's approach to catastrophic AI risk.
Khanna argues that leaders in both Washington and Beijing are not treating AI risks with enough urgency. He called a recent Trump-Xi meeting on AI constructive but inadequate, and dismissed an upcoming White House session with tech executives as a branding exercise. In his view, companies with profit motives should not be the ones writing safety rules, and politicians lack the technical expertise to lead the debate.
The congressman ultimately wants a US-China treaty banning recursive self-improvement and monitoring frontier labs, though he acknowledges that is unlikely soon. He also expects the Chinese companies may not respond fully, saying his goal is to shame them and spark a debate. Skepticism remains on the US side as well: Republican leaders have doubted China would honor such an agreement, and President Trump has said he wants to leave superintelligence where it is.