Representative Ro Khanna (D-CA) called for a treaty between the United States and China to prevent artificial intelligence from causing global disruption. The Congressman specifically requested that the Office of the Director of National Intelligence (ODNI) assess the US response to a scenario where an AI laboratory loses control of its bots.
Furthermore, Khanna asked the ODNI to evaluate the Chinese government's methods for assessing risks associated with catastrophic AI developments. In letters sent to DeepSeek, Alibaba, and Moonshot AI, Khanna sought documentation detailing efforts to reach 'superintelligence' and recursive self-improvement (RSI).
He also inquired about implementing safeguards and 'kill switches' for the technology. Khanna's advocacy follows a recent meeting between Chinese President Xi Jinping and US President Donald Trump at the White House last week.
In an interview on Monday, Khanna told The Verge that the meeting between Trump and Xi Jinping was constructive but not adequate. Khanna stated that US companies maximizing profits should not be the ones setting the rules for safety.
He expressed a desire for a US-China treaty that would ban RSI and monitor frontier AI laboratories. Other lawmakers also voiced similar concerns, with Senator Bernie Sanders (I-VT) calling for an AI treaty between the US and China.
However, John Moolenaar (R-MI) expressed doubt regarding the Chinese government's commitment to upholding any terms of an AI treaty. Leading US AI companies have called for the US to coordinate with China to slow the rate of development.
Khanna noted that the Chinese AI companies failed to reply to an earlier letter requesting a commitment to slow AI development alongside OpenAI and Anthropic. Khanna stated that his objective is to 'shame them' and 'spark a debate' among Chinese AI companies regarding safety.
Despite these concerns, Khanna remains confident that the US will set the standards for AI safety because many Chinese models draw from US models.
