Khanna Pushes for US-China AI Treaty Amid Growing Risks

Overview

As President Donald Trump prepares to meet with technology and AI chief executives in Washington, Rep. Ro Khanna (D-CA) is intensifying his call for a formal US-China AI treaty. The top Democrat on the House Select Committee on China argues that such an agreement is essential to prevent the technology from causing catastrophic harm globally. However, Khanna acknowledges that securing action from leaders in both nations remains a significant challenge.

Khanna contends that neither American nor Chinese leadership currently demonstrates the necessary urgency to rein in the risks posed by advanced artificial intelligence. His push follows recent high-level diplomatic engagements, including a meeting between President Trump and Chinese President Xi Jinping, which he describes as constructive but insufficient on its own.

Legislative Push and Intelligence Requests

Khanna has initiated a two-pronged legislative strategy involving official correspondence with US intelligence agencies and major Chinese AI developers. In letters shared exclusively with reporting outlets, he requested that the Office of the Director of National Intelligence (ODNI) assess how effectively the United States could respond to a scenario where an AI laboratory loses control of its autonomous agents.

"American and Chinese interests could not be more different on many issues," Khanna writes in his correspondence to Director of National Intelligence Jay Clayton. "But on this discrete concern, it benefits neither country for humanity to lose control of our destiny."

The requested assessment aims to clarify potential response protocols for incidents similar to the recent hack involving OpenAI agents and Hugging Face. Additionally, Khanna asked the ODNI to evaluate the Chinese government’s current approach to assessing catastrophic AI risks, seeking information that could aid Congress and the White House in crafting a bilateral agreement.

Inquiry into Chinese Developers

Simultaneously, Khanna directed letters to prominent Chinese AI companies, including DeepSeek, Alibaba, and Moonshot AI. These firms are all actively pursuing high-level artificial intelligence capabilities. The correspondence requested documentation regarding their efforts to achieve superintelligence and recursive self-improvement (RSI), a critical milestone where AI systems can iteratively enhance their own code and architecture.

The Representative also asked these companies to disclose:

• Any existing safeguards or "kill switches" implemented to halt runaway AI processes.
• Their willingness to commit to inspections by non-governmental agencies should a treaty be established.
• Their stance on international monitoring frameworks for frontier AI labs.

Neither the ODNI nor the mentioned Chinese companies immediately responded to requests for comment regarding these inquiries.

Skepticism Toward Industry Self-Regulation

Khanna’s advocacy extends beyond international diplomacy to a critique of domestic industry practices. He expressed strong skepticism regarding a planned meeting between the White House, House Speaker Mike Johnson, and several American tech executives. Khanna characterized the event as primarily a branding exercise designed to reassure the public that companies are acting responsibly.

He argued that entrusting profit-driven corporations with the authority to define safety standards is fundamentally flawed. To illustrate his point, he drew a parallel to the pharmaceutical industry, questioning whether society would trust drug executives to determine the safety protocols for life-saving medications without independent oversight.

"Why should we trust companies that have a motive to maximize profits to be making the rules for what is safe?" Khanna stated.

While he did not label the executives as "bad people," he emphasized that structural incentives within the tech industry often conflict with rigorous safety implementation. He noted that while US firms are increasingly engaging in safety conversations, the primary driver remains competitive rather than purely ethical.

Political Landscape and Treaty Feasibility

The push for an AI treaty faces considerable political headwinds. President Trump recently posted on Truth Social that he intended to keep AI regulation exactly where it is, aligning with what he perceived as China’s position. House Speaker Johnson echoed this sentiment, warning against regulations that might stifle innovation or cause the United States to lose the AI race to China.

Despite this resistance, notable figures across the spectrum are calling for action. Senator Bernie Sanders has similarly advocated for a US-China treaty, while leading US AI companies have publicly urged coordination with Beijing to slow the pace of development. Conversely, Republicans like House China Committee Chair John Moolenaar remain skeptical that the Chinese government would honor the terms of any agreement.

Potential Compromises

Recognizing that a comprehensive ban on recursive self-improvement is unlikely to pass immediately, Khanna is proposing the formation of a bilateral working group. This body would consist of experts from both nations tasked with suggesting safety recommendations and developing methods for ensuring compliance.

Khanna remains confident that the United States will ultimately drive global AI safety standards. He cites the widespread reliance of Chinese models on distilled versions of US technology as a factor that gives American experts significant leverage. He believes that while politicians may be out of touch with the technical realities, the scientists and engineers on both sides of the Pacific are well aware of the existential risks involved.

"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 a technical expertise."

Ultimately, Khanna hopes his letters serve to shame reluctant actors and spark internal debate among Chinese AI companies about safety priorities. Although full transparency from Chinese firms is unlikely, he aims to establish a record of diplomatic effort and highlight the urgent need for international cooperation in managing the future of artificial intelligence.