Representative Ro Khanna (D-CA), a senior Democrat on the House Select Committee on China, is calling for a treaty between the United States and China to prevent artificial intelligence from causing global harm. In letters shared with The Verge, Khanna urged US intelligence agencies and major Chinese AI companies—including DeepSeek, Alibaba, and Moonshot AI—to address potential risks associated with advanced AI systems.
Khanna's proposal includes a request for the Office of the Director of National Intelligence (ODNI) to assess how the US could respond to scenarios where an AI lab loses control of its agents. The lawmaker also sought documentation from Chinese companies regarding their efforts toward achieving "superintelligence" and recursive self-improvement (RSI)—a process where AI systems can autonomously improve their own capabilities.
The letters specifically ask whether these companies would be willing to implement safeguards, such as "kill switches," and if they would commit to inspections by non-governmental agencies if a treaty were established. Khanna expressed skepticism regarding the ability of tech executives to self-regulate, comparing the situation to the pharmaceutical industry.
While Khanna's call for a ban on RSI and increased monitoring of frontier AI labs faces political hurdles, he advocated for the creation of a working group of experts from both nations to develop safety recommendations and compliance methods.
Sources
- Will Chinese AI companies slow down? A top House Democrat wants answers （The Verge AI, 2026-09-29）