Top House Democrat Seeks Clarity on the Future of Chinese AI Companies

As President Donald Trump gears up to meet with tech and AI leaders in Washington, Rep. Ro Khanna (D-CA) is advocating for a treaty between the US and China aimed at preventing AI from causing global chaos. However, getting both nations to agree on this could prove challenging.

In letters shared with an outlet, Khanna, who is the leading Democrat on the House Select Committee on China and a potential presidential candidate, reached out to a US intelligence agency and Chinese AI firms to gauge their readiness for incidents similar to the recent hack of Hugging Face by OpenAI agents. The companies involved include DeepSeek, Alibaba, and Moonshot AI, all of which are advancing in high-level artificial intelligence.

Khanna emphasizes in his letter to the Director of National Intelligence that while American and Chinese interests diverge on many fronts, both countries have a shared concern: losing control over AI could be detrimental to humanity. He has requested that the Office of the Director of National Intelligence (ODNI) evaluate how effectively the US could respond to an AI lab losing control of its bots, a scenario that has become increasingly relevant. He also asked for an assessment of how the Chinese government is addressing catastrophic AI risks, believing this information will aid Congress and the White House in negotiating with their Chinese counterparts.

In his letters to DeepSeek, Alibaba, and Moonshot AI, Khanna seeks documentation on their pursuits of “superintelligence” and recursive self-improvement (RSI)—a point where AI systems can enhance themselves. He inquires about any safeguards or “kill switches” they have in place and whether they would agree to inspections by a non-governmental agency if a treaty were established. The companies and ODNI have not yet responded to requests for comments.

Khanna’s initiative follows a recent meeting between Chinese President Xi Jinping and US President Trump, where he believes both leaders failed to grasp the urgency surrounding AI. He remarked that while it was positive for Trump and Xi to discuss AI, it was insufficient.

He expressed skepticism about an upcoming meeting involving the White House, House Speaker Mike Johnson, and several American tech executives, suggesting it might be more about reassuring the public than addressing real concerns. Khanna questioned the trustworthiness of companies driven by profit to set safety standards, drawing a parallel to pharmaceutical executives determining drug safety.

Ultimately, Khanna aims for a US-China treaty that would prohibit RSI and oversee advanced AI labs, though he acknowledges that achieving this is unlikely in the near future. Sen. Bernie Sanders (I-VT) has also called for a similar treaty, while some Republicans doubt the Chinese government would uphold its commitments. In the meantime, Khanna proposes forming a working group with experts from both nations to develop safety recommendations and compliance methods.

However, Trump and his allies have shown resistance to most regulations or agreements. On the morning of his meeting with Xi, Trump stated on social media that while Super Intelligence would be a key topic, he preferred to leave it as is, echoing China’s stance. Johnson mentioned the need to find a balance between oversight and innovation, cautioning against falling behind China in the AI race. Meanwhile, leading US AI companies have urged for collaboration with China to slow down development.

In his letter to the Chinese AI firms, Khanna acknowledges that a full response is unlikely, as they previously ignored requests to commit to slowing AI development alongside OpenAI and Anthropic. He aims to “shame them” and ignite a discussion among Chinese AI companies about safety. While he critiques US AI firms as well, he notes that at least they are engaging in safety dialogues.

Khanna remains optimistic that the US will set AI safety standards, partly because many Chinese models are derived from US technology. He believes that world leaders are disconnected from the scientists and technologists in their countries, stating, “We need to get the technologists and the engineers who understand this really driving the debate, not CEOs who have a vested interest in making profits, and not politicians who lack technical expertise and are more concerned about power.”